A

Project Stage-II Report

on
"SmartHire:AI Interview Platform"
Submitted By
Patil Dipak Rajendra (211106022)
More Mohit Mangolal (211106038)
Warke Gunjan Chandrakant (211106046)
Patil Gaurav Jitendra (211106065)
Under the Guidance of
Prof. Dr. U. M. Patil

The Shirpur Education Society’s

Department of Computer Science & Engineering (Data

Science)
R. C. Patel Institute of Technology Shirpur - 425405.
Maharashtra, India
[2024-25]
 A
Project Stage-II Report
on
"SmartHire:AI Interview Platform"
Submitted By
Patil Dipak Rajendra (211106022)
More Mohit Mangolal (211106038)
Warke Gunjan Chandrakant (211106046)
Patil Gaurav Jitendra (211106065)
Under the Guidance of
Prof. Dr. U. M. Patil

The Shirpur Education Society’s

Department of Computer Science & Engineering (Data

Science)
R. C. Patel Institute of Technology Shirpur - 425405.
Maharashtra, India
[2024-25]
 The Shirpur Education Society’s
R. C. Patel Institute of Technology
Shirpur, Dist. Dhule (M.S.)
Department of Computer Science & Engineering (Data Science)

CERTIFICATE

This is to certify that the Project Stage-II entitled “SmartHire:AI Interview Plat-
form” has been carried out by the team:

Dipak Ranjendra Patil

Mohit Mangolal More
Gunjan Chandrakant Warke Gaurav Jintendra Patil

under the guidance of Prof.Dr.U.M.Patil in partial fulfillment of the requirement for

the degree of Bachelor of Technology in Department of Computer Science & Engineering
(Data Science) (Semester-VIII) of Dr. Babasaheb Ambedkar Technological University,
Lonere during the academic year 2024-25.

Date:
Place: Shirpur

Guide Project Coordinator

Prof. Dr. U. M. Patil Prof. Dr. P. S. Sanjekar

H. O. D. Director
Prof. Dr. U. M. Patil Prof. Dr. J. B. Patil
 ACKNOWLEDGEMENT
It is our privilege to express our sincerest regards to our project guide Prof.
Dr. U. M. Patil for their valuable inputs, adorable guidance, encourage-
ment, whole-hearted cooperation, and constructive criticism throughout the
duration of our project.

We deeply express our sincere thanks to our Head of Department Prof.

Dr. U. M. Patil for encouraging and allowing us to present the project on
the topic “SmartHire: AI Interview Platform” at our department premises
for the partial fulfilment of the requirements leading to the award of a B-Tech
degree.

We are also grateful to Director Prof. Dr. J. B. Patil, for his kind
support and guidance. Last but not the least, we express our sincere thanks
to all our friends who have patiently extended all sorts of help for accom-
plishing this undertaking.
Thank you for your kind support.

Project Team:
Patil Dipak Ranendra [PRN No.211106022]
More Mohit Mangolal [PRN No.211106041]
Warke Gunjan Chandrakant[PRN No.211106046]
Patil Gaurav Jitendra [PRN No.211106065]
Contents

1 Introduction 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Problem statement . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2 Literature Survey 10
2.1 Review of existing system . . . . . . . . . . . . . . . . . . . . 10
2.2 Limitations of existing system . . . . . . . . . . . . . . . . . . 11
2.3 Related concept . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3 Proposed System 15
3.1 Working . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2 Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3 Hardware and software requirement . . . . . . . . . . . . . . . 21

4 Methodology 23
4.1 System Design . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

5 Implementation Details 29
5.1 Login Page Interface . . . . . . . . . . . . . . . . . . . . . . . 29
5.2 Dashboard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.3 Interview Instructions . . . . . . . . . . . . . . . . . . . . . . 31
5.4 Experience Level Selection . . . . . . . . . . . . . . . . . . . . 33
5.5 Permission Check . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.6 Interview Interface . . . . . . . . . . . . . . . . . . . . . . . . 35
5.7 Feedback After Interview . . . . . . . . . . . . . . . . . . . . . 37

6 Testing 41
6.1 Testing Phase 1 . . . . . . . . . . . . . . . . . . . . . . . . . . 41
6.2 Testing Phase 2 . . . . . . . . . . . . . . . . . . . . . . . . . . 42

1
7 Results and Discussion 44
7.1 Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
7.2 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

8 Conclusion 47

BIBLIOGRAPHY 48

2
List Of Figures

3.1 Block Diagram of the Proposed System . . . . . . . . . . . . . 16

3.2 Workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.1 System Architecture . . . . . . . . . . . . . . . . . . . . . . . 24
5.1 Login Interface . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.2 Dashboard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.3 Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.4 Different Levels Of Interview . . . . . . . . . . . . . . . . . . . 34
5.5 Permissions For Camera,Audio,etc. . . . . . . . . . . . . . . . 35
5.6 Interview Interface . . . . . . . . . . . . . . . . . . . . . . . . 36
5.7 Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

3
List Of Tables

Page No.
3.3 Hardware and Software Requirement . . . . . . . . . . . . . . 21

4
 Abstract
The Interview preparation is a critical aspect of career development, yet many candidates
lack access to effective and personalized practice tools. To address this gap, we have
developed an AI-powered mock interview platform that simu- lates real-world interview
experiences using advanced natural language processing (NLP) and other technology. The
platform conducts automated interviews across various domains, evaluates user responses
in real time, and provides detailed, constructive feedback on aspects such as content
quality, communication skills, confidence, and body language (if applicable). By offering
a scalable, accessible, and unbiased practices environment, our solution empowers users
to improve their performance, reduce interview anxiety, and build confidence—ultimately
enhancing their chances of success in actual interviews.

1
Chapter 1

Introduction

In an era defined by rapid technological advancement and an increasingly competitive job

market, traditional methods of interview preparation are no longer sufficient. With the
growing demand for skilled professionals across various sectors, particularly in technology-
driven industries, it has become essential for candidates to not only possess technical
knowledge but also demonstrate strong communication, critical thinking, and problem-
solving skills during job interviews. To address this need, the AI Mock Interview Platform
emerges as a powerful and innovative solution designed to transform the way individuals
prepare for interviews.
The AI Mock Interview Platform leverages the capabilities of artificial intelligence,
natural language processing (NLP), and machine learning (ML) to simulate real-time,
interactive interview experiences. Unlike conventional mock interviews that rely on hu-
man interviewers or static question sets, this platform offers dynamic, intelligent, and
personalized sessions tailored to the candidate’s selected job role, domain expertise, and
experience level. This ensures a highly realistic and adaptive interview environment that
mirrors actual recruitment processes.
One of the core features of the platform is its ability to generate relevant and context-
specific interview questions. Whether the candidate is applying for a role in software
engineering, data science, finance, marketing, or human resources, the platform analyzes
their profile and job description to formulate questions that assess both technical and
behavioral competencies. These questions can range from coding challenges and case
studies to situational and HR-based queries, providing comprehensive coverage of the
interview spectrum.

1
 R.C.P.I.T.,Shirpur

After each session, the platform evaluates the candidate’s responses using AI mod-
els trained on large datasets of successful interview patterns and industry standards.
Feedback is provided instantly, covering various aspects such as content accuracy, log-
ical reasoning, communication skills, body language (if video input is used), and time
management. This immediate, detailed feedback loop allows users to identify areas of
improvement and track their progress over time, making the platform a valuable self-
assessment tool.
Beyond individual interviews, the AI Mock Interview Platform also includes features
for peer-to-peer practice and collaborative learning. Users can participate in commu-
nity mock interviews where they can role-play as both interviewer and interviewee. This
encourages knowledge sharing, confidence building, and exposure to a variety of ques-
tion styles and interview techniques. Additionally, users benefit from curated learning
resources such as video tutorials, sample answers, industry-specific guides, and tips from
recruiters, all integrated into the platform for a holistic learning experience.
Accessibility and flexibility are central to the platform’s design. As a cloud-based
service, it can be accessed from any device with an internet connection, enabling users
to practice interviews anytime and anywhere. This is particularly beneficial for students,
working professionals, and job seekers in remote or underserved areas who may lack access
to formal career services. With support for multiple languages and regional accents, the
platform ensures inclusivity and global reach.[1]
Another notable advantage of using an AI-driven platform is its scalability and consis-
tency. While human-led mock interviews may be subject to bias or variability in feedback
quality, AI ensures uniform evaluation across users. It also allows institutions such as
universities, training centers, and corporate learning departments to deploy the platform
at scale, providing thousands of learners with standardized and personalized interview
preparation support.
Furthermore, the AI Mock Interview Platform incorporates analytics and reporting
features that help users visualize their performance trends. These insights can be par-
ticularly useful for career coaches, placement officers, or hiring managers who want to
monitor a candidate’s readiness or progress. It also motivates users to set goals, measure
improvement, and stay engaged in their learning journey.
As technology continues to reshape every aspect of our lives, from communication to

SmartHire: AI Mock Interview Platform 2

 R.C.P.I.T.,Shirpur

education and employment, AI-powered tools like this platform represent a significant
leap forward in professional development. They not only democratize access to quality
interview preparation but also empower individuals to take ownership of their career
growth in a data-driven and efficient manner.
In conclusion, the AI Mock Interview Platform is more than just a simulation tool—it
is an intelligent career mentor that combines cutting-edge technology with practical train-
ing to prepare candidates for success. By offering realistic, interactive, and insightful
interview experiences, it equips users with the confidence and competence they need to
excel in competitive job markets. As organizations continue to value talent that can
adapt, communicate, and innovate, platforms like these will play an increasingly vital
role in shaping the workforce of the future.

1.1 Background
The evolution of recruitment practices has undergone a significant transformation in re-
cent years, driven by the dual forces of globalization and rapid technological advancement.
The modern job market is no longer confined to local or regional talent pools; instead,
it operates within a global framework where candidates must compete with peers world-
wide. This increased competition, combined with the growing complexity of roles across
industries, has heightened the expectations placed on job seekers. Employers now seek
individuals who not only possess domain-specific expertise but also demonstrate strong
communication, problem-solving abilities, and adaptability—traits that are often assessed
during job interviews.
Traditionally, interview preparation has been a largely informal process. Candidates
have typically relied on career counselors, mentors, peers, or publicly available resources
such as books, blogs, and forums to prepare for interviews. While these sources pro-
vide helpful guidance, they often lack personalization, structure, and real-time feed-
back—factors that are essential for meaningful skill development. For candidates without
access to professional coaching or extensive networks, particularly those from underserved
or remote regions, preparing effectively for interviews can be a daunting challenge.
In response to these limitations, the field of recruitment and career development has
increasingly turned to digital solutions. Among the most significant innovations in this

SmartHire: AI Mock Interview Platform 3

 R.C.P.I.T.,Shirpur

domain is the application of Artificial Intelligence (AI) to simulate interview scenarios.

Over the past decade, AI has made substantial inroads into various stages of the hiring
lifecycle—from automated resume screening and initial candidate engagement through
chatbots to predictive analytics used in candidate evaluation. The integration of AI
into interview preparation represents a natural extension of this trend, leading to the
development of AI Mock Interview Platforms.[2]
AI Mock Interview Platforms are designed to simulate real-life interview experiences
using advanced technologies such as Natural Language Processing (NLP), Machine Learn-
ing (ML), and, in some cases, Computer Vision. NLP enables these platforms to interpret
and respond to human language in a conversational manner, while ML allows the system
to learn from vast datasets of interview interactions and improve over time. Computer
Vision, when incorporated, allows for the analysis of non-verbal cues in video interviews,
such as eye movement, facial expressions, and gestures—factors that are critical in real
interviews but often overlooked in conventional preparation methods.
Early versions of these platforms were relatively limited, often functioning as sim-
ple QA interfaces or scripted chatbots. However, advancements in AI, especially deep
learning and contextual understanding, have led to more sophisticated systems capable
of conducting adaptive interviews. These platforms can now tailor questions based on
user responses, provide follow-up questions that mimic human interviewer behavior, and
deliver detailed performance analytics. This shift has transformed AI mock interviews
from basic practice tools into comprehensive learning environments.
Another critical aspect of these platforms is their role in democratizing access to
interview preparation. By offering scalable, affordable, and on-demand resources, AI
Mock Interview Platforms help bridge gaps caused by socioeconomic, geographic, or in-
stitutional disparities. This is particularly important in developing regions, where access
to high-quality career support services may be limited or nonexistent. Through a web
or mobile interface, any candidate with internet access can simulate interviews, receive
feedback, and build confidence.
These platforms are not only beneficial for individual users but also hold value for
educational institutions and employers. Universities and training academies can integrate
AI interview systems into their career services, offering students a standardized method
of preparing for placements and internships. Employers, on the other hand, can use these

SmartHire: AI Mock Interview Platform 4

 R.C.P.I.T.,Shirpur

tools for internal training, workforce reskilling, and even preliminary candidate screening.
In summary, the emergence of AI Mock Interview Platforms represents a significant
milestone in the intersection of technology and career development. By addressing the
shortcomings of traditional preparation methods and leveraging the power of AI, these
platforms provide a structured, efficient, and inclusive approach to interview readiness.
As AI continues to evolve and permeate more aspects of professional development, its
role in shaping the future of hiring and workforce preparation will only become more
pronounced.

1.2 Motivation:
Securing employment in today’s fast-paced, competitive job market presents a consider-
able challenge for job seekers. While academic qualifications and technical skills remain
critical, the ability to perform well in job interviews often serves as the decisive factor
in the hiring process. However, despite the high stakes of interviews, many candidates
find themselves underprepared due to a lack of personalized training, real-time feedback,
or exposure to realistic interview scenarios. This gap between theoretical knowledge
and practical readiness forms the central motivation for the development of an AI Mock
Interview Platform.
One of the primary challenges faced by candidates is the lack of access to consistent
and high-quality interview coaching. Professional interview preparation services are often
expensive and geographically limited, making them inaccessible to many students, fresh
graduates, or career switchers. Additionally, individuals from non-urban or underserved
regions often do not have access to experienced mentors or structured career support
systems. This inequality can result in significant disparities in career outcomes, even
among equally qualified candidates. The motivation to create a more equitable, scalable,
and accessible solution stems from the desire to bridge this gap and democratize career
development resources.
Moreover, traditional methods of interview preparation—such as mock sessions con-
ducted by faculty, peers, or mentors—are often inconsistent and subjective. These ses-
sions can be limited by human bias, time constraints, and lack of domain expertise. In
contrast, an AI-powered platform can offer objective, data-driven feedback, available on-

SmartHire: AI Mock Interview Platform 5

 R.C.P.I.T.,Shirpur

demand, and personalized to the user’s field and experience level. The use of AI also
allows for repeatable, adaptive practice, which enhances learning outcomes and boosts
candidate confidence.
The rapid evolution of recruitment practices, especially with the rise of remote hiring
and AI-driven applicant screening, further emphasizes the need for candidates to adapt
and stay prepared. Interviews are increasingly conducted virtually, requiring not only
good content delivery but also strong online communication skills. This shift motivated
the integration of video-based simulations and non-verbal analysis into the platform,
allowing users to practice and refine these essential soft skills.[3]
From an institutional perspective, there is also a growing demand for scalable inter-
view training tools that can support large cohorts of students or employees. Colleges,
universities, and training centers often struggle to provide one-on-one coaching to all
learners. The motivation to implement a centralized AI platform enables institutions to
deliver standardized, effective, and scalable interview practice sessions, improving overall
placement success rates.
In essence, the development of the AI Mock Interview Platform is driven by the vision
to empower job seekers with a fair and effective way to prepare for interviews—anytime,
anywhere. It is inspired by the need to modernize interview preparation, reduce barriers
to career growth, and enhance the employability of candidates across all backgrounds.
By combining technology with practical training, the platform seeks to instill confidence,
provide constructive feedback, and ultimately increase success rates in the job selection
process.

1.3 Problem Statement

The process of securing a job is often heavily influenced by how well a candidate performs
during the interview stage. Regardless of qualifications or technical competence, the
ability to effectively communicate, respond confidently, and demonstrate problem-solving
skills during an interview frequently becomes the deciding factor for selection. However,
a large number of candidates, particularly fresh graduates and career switchers, struggle
with interview preparation due to a lack of structured practice opportunities and access
to high-quality feedback.

SmartHire: AI Mock Interview Platform 6

 R.C.P.I.T.,Shirpur

Traditional methods of interview preparation—such as mock interviews conducted

by faculty, peers, or career counselors—tend to be inconsistent, time-consuming, and
highly dependent on the availability and expertise of the evaluator. These approaches
are also limited by subjective biases, lack of standardization, and an inability to offer
in-depth, real-time performance analytics. Additionally, the reach of such methods is
often constrained within academic institutions or urban centers, leaving candidates from
remote or underserved areas with little to no guidance.
With the rise of remote work, digital hiring practices, and the use of automated
systems in recruitment, the landscape of job interviews has undergone a major transfor-
mation. Virtual interviews, AI-powered assessments, and competency-based evaluations
have become commonplace, requiring candidates to adapt to new formats. Yet, most
preparation tools do not reflect these changes and fail to prepare users for the nuances
of digital interviews—such as maintaining eye contact with a webcam, managing screen
presence, or handling behavioral questions from automated systems.
Compounding the problem, educational institutions and training centers are under
increasing pressure to improve student placement rates but often lack the resources to
provide individual interview coaching at scale. Faculty members and placement cells may
not have the bandwidth or industry-specific expertise to conduct detailed, personalized
mock interviews for every student. This results in inconsistent levels of preparedness
among candidates, which directly affects their employability.
Moreover, candidates often lack awareness of their own communication gaps or be-
havioral tendencies due to the absence of objective, data-backed feedback. This can
lead to repeated mistakes in interviews, reduced self-confidence, and missed opportuni-
ties—despite possessing strong academic or technical credentials.
Therefore, there exists a critical need for an intelligent, accessible, and scalable solu-
tion that can overcome these challenges. The problem lies in the absence of a platform
that not only simulates realistic interview scenarios but also provides personalized, ac-
tionable feedback using advanced technologies. An AI Mock Interview Platform, driven
by Natural Language Processing (NLP), Machine Learning (ML), and Computer Vision,
can fill this gap by offering a standardized, immersive, and adaptive interview practice ex-
perience. Such a solution would empower users to develop interview skills independently,
gain insights into their performance, and improve their confidence—thereby increasing

SmartHire: AI Mock Interview Platform 7

 R.C.P.I.T.,Shirpur

their chances of success in real-world job interviews.

1.4 Objectives
The objectives of this project are multifaceted and aim to achieve several key out-
comes:
• To Engineer a Highly Realistic and Immersive Interview Simulation: The platform
should be capable of replicating the key aspects of real job interviews, including the
format (e.g., one-on-one, panel), the pacing of questions, and the types of questions
commonly encountered (e.g., behavioral, technical, situational, case study). This
involves sophisticated algorithms that can adapt the flow of the interview based on
the candidate’s responses and introduce realistic challenges.

• To Deliver Granular, Personalized, and Actionable Feedback Across Multiple Di-

mensions: The AI should analyze various facets of a candidate’s performance, in-
cluding the content and structure of their answers, their verbal communication style
(clarity, conciseness, tone, pacing, grammar, vocabulary, use of filler words), and
potentially their non-verbal cues (eye contact, posture, facial expressions) if video
analysis is integrated. This feedback should be specific, linked to concrete examples
from the interview, and provide clear recommendations for improvement.

• To Offer Highly Tailored Interview Practice Aligned with Specific Job Roles, Indus-
tries, and Company Requirements: The platform should allow users to customize
their mock interviews by selecting specific job titles, industries, and even potentially
inputting company-specific information. The AI should then generate relevant ques-
tions and evaluate responses based on the typical expectations and requirements of
those contexts.

• To Significantly Bolster Candidate Confidence and Alleviate Interview-Related Anx-

iety: By providing a safe and repeatable practice environment, coupled with con-
structive feedback, the platform aims to build candidates’ familiarity with the in-
terview process, reduce their fear of the unknown, and increase their self-assurance
in their ability to perform well.

SmartHire: AI Mock Interview Platform 8

 R.C.P.I.T.,Shirpur

• To Hone Candidates’ Verbal Communication and Articulation Skills: Through re-

peated practice and targeted feedback on their responses, users can learn to express
their experiences, skills, and qualifications more clearly, concisely, and persuasively.
The AI can help identify areas where candidates struggle with clarity, organization,
or the effective use of storytelling techniques.

• To Equip Candidates with Insights into Common Interview Questions and Effective
Response Strategies: The platform should expose users to a wide range of frequently
asked interview questions and provide guidance on effective frameworks and tech-
niques for answering them, such as the STAR method for behavioral questions or
structured approaches for case studies. The AI can also analyze the quality and
completeness of the candidate’s responses against ideal answers.

• To Establish a Scalable and Universally Accessible Solution for Interview Prepa-

ration:The AI-powered nature of the platform ensures that it can cater to a large
number of users simultaneously, overcoming the logistical constraints of human-
led coaching. Its online accessibility breaks down geographical barriers and offers
24/7 availability, making high-quality interview practice available to anyone with
an internet connection.

• To Furnish Job Seekers with Data-Driven Insights into Their Interview Proficiency
and Progress:The platform should track key performance metrics across multiple
practice sessions, allowing users to monitor their improvement over time and iden-
tify persistent areas of weakness that require further attention. This data-driven
approach empowers candidates to take ownership of their development and focus
their efforts effectively.

Overall, the objective of this project is to create a secure, decentralized, and user-friendly
digital locker system using blockchain technology and Solidity smart contracts, with the
aim of enhancing data security and privacy for individuals and organizations.

SmartHire: AI Mock Interview Platform 9

Chapter 2

Literature Review

2.1 Review of Existing System

• Human-Led Mock Interview Services:These represent the traditional approach, in-
volving career counselors, mentors, or HR professionals who conduct mock inter-
views and provide feedback. Services range from university career centers to private
coaching firms. These systems offer personalized interaction and nuanced feedback
based on human experience and intuition. They can often tailor the interview to
specific roles and industries based on the interviewer’s expertise. Platforms also
exist that connect job seekers with experienced professionals for paid or pro-bono
mock interviews, leveraging online video conferencing tools for remote sessions.

• Basic Online Question Banks and Practice Platforms:Numerous websites and plat-
forms offer extensive lists of common interview questions categorized by type (be-
havioral, technical, situational). Some provide basic frameworks or suggested ap-
proaches for answering these questions. These platforms primarily focus on content
preparation and self-assessment. They may include features like recording practice
answers for self-review or comparing one’s responses to sample answers. Examples
include Glassdoor’s interview question database and various career advice websites.

• Early-Stage Automated Mock Interview Tools:With the advancement of AI, early

attempts at automated mock interview systems emerged. These often involved
text-based or simple voice-based interactions with rule-based systems. They could

10
 R.C.P.I.T.,Shirpur

pose questions and provide rudimentary feedback based on keyword recognition or

pre-programmed logic. Some tools focused on specific aspects like elevator pitch
practice or answering common behavioral questions. These systems offered the
advantage of 24/7 availability but lacked the sophistication for nuanced interaction
and comprehensive feedback.

• Avatar-Based Virtual Interview Simulations:Some platforms utilize virtual avatars

to conduct mock interviews in a simulated 3D environment. These systems aim
to enhance the realism of the experience by incorporating visual cues and allowing
candidates to practice their non-verbal communication. The feedback mechanisms
in these systems vary in sophistication, ranging from basic assessments of pacing
and pauses to more advanced analysis of body language and facial expressions.
Development in this area often intersects with virtual reality (VR) and augmented
reality (AR) technologies to create more immersive experiences.[4]

• AI-Powered Mock Interview Platforms (Emerging Generation):This represents the

most recent evolution, leveraging advancements in NLP, ML, and computer vision.
These platforms aim to provide a more realistic and comprehensive mock interview
experience. They can analyze the content of responses, the candidate’s commu-
nication style , and potentially non-verbal cues. Feedback is often data-driven,
identifying specific areas for improvement based on patterns learned from large
datasets of successful and unsuccessful interview performances. Examples include
platforms that offer scoring on various aspects of the interview, suggest alternative
phrasing, and provide insights into potential areas of weakness.[5]

2.2 Limitations of Existing System

Despite the progress in interview preparation tools, existing systems still exhibit sev-
eral limitations
1 Limitations of Human-Led Systems: While offering valuable personalized feedback,
human-led mock interviews suffer from scalability issues, scheduling constraints, and
potential inconsistencies in feedback quality due to varying levels of interviewer ex-
perience and biases. They can also be expensive, limiting accessibility for many job

SmartHire: AI Mock Interview Platform 11

 R.C.P.I.T.,Shirpur

seekers. Furthermore, replicating the specific nuances of a wide range of technical

roles or niche industries can be challenging even for experienced professionals.

2 Limitations of Basic Online Question Banks and Practice Platforms:These systems

primarily focus on content preparation and lack the interactive and feedback-driven
elements crucial for improving interview delivery skills. Self-assessment can be sub-
jective and may not accurately identify weaknesses in communication style or overall
presentation. The lack of a simulated interview environment does not adequately
prepare candidates for the pressure and flow of a real interview.

3 Limitations of Basic Online Question Banks and Practice Platforms: These systems
primarily focus on content preparation and lack the interactive and feedback-driven
elements crucial for improving interview delivery skills. Self-assessment can be sub-
jective and may not accurately identify weaknesses in communication style or overall
presentation. The lack of a simulated interview environment does not adequately
prepare candidates for the pressure and flow of a real interview.

4 Limitations of Early-Stage Automated Mock Interview Tools: These systems of-

ten lack the natural language understanding capabilities to engage in meaningful
back-and-forth conversation or to comprehend the subtleties of human communica-
tion. Feedback is typically superficial and based on keyword matching rather than
a deeper understanding of the context and quality of the response. They often
struggle with open-ended questions and providing nuanced guidance.

5 Limitations of Avatar-Based Virtual Interview Simulations: While improving the

realism of the visual environment, the effectiveness of avatar-based systems heav-
ily relies on the sophistication of their AI-powered feedback mechanisms. If the
underlying analysis of verbal and non-verbal cues is rudimentary, the added visual
dimension may not translate into significantly better preparation. Technical limi-
tations in accurately capturing and interpreting subtle human expressions can also
hinder the quality of feedback. The cost and accessibility of VR/AR-based systems
can also be a barrier.

6 Limitations of Emerging AI-Powered Mock Interview Platforms: Even the most

advanced AI platforms still face challenges. Accurately replicating the complexity

SmartHire: AI Mock Interview Platform 12

 R.C.P.I.T.,Shirpur

of human interaction, including understanding intent, humor, and cultural nuances,

remains a significant hurdle. Current AI models may sometimes provide generic or
superficial feedback, particularly in response to highly nuanced or unconventional
answers. The ethical considerations surrounding data privacy and potential biases
in the training data also need careful consideration. Furthermore, the ability to
truly assess complex problem-solving skills or evaluate ”fit” within a specific com-
pany culture remains a challenge for purely AI-driven systems. The lack of genuine
human empathy and the inability to provide motivational support in the same way
a human mentor can also be considered limitations.

7 Data Silos: The existing system may create data silos, where data is stored in
separate locations or formats, making it difficult to access and manage information
efficiently.

2.3 Related Concept

AI interview platforms like SmartHire leverage advancements in artificial intelligence,
natural language processing (NLP), speech recognition, and machine learning to create
intelligent, interactive environments for mock interviews and career readiness. Natural
Language Processing (NLP) plays a pivotal role in understanding and evaluating user
responses in both text and speech format. NLP enables the system to analyze grammar,
context, tone, and relevance of answers during an interview, offering precise feedback to
the candidate. Machine Learning (ML) is used to continuously improve the accuracy
of candidate evaluation. By training models on a wide range of interview datasets, the
system learns to identify successful patterns in responses and provide recommendations
on how users can enhance their performance. Speech Recognition technology allows the
platform to analyze spoken responses in real time. Features like speech fluency, clar-
ity, tone, and pitch are assessed to simulate a realistic interview environment and help
users identify areas for improvement. Sentiment Analysis and Emotion Detection help
gauge a candidate’s confidence level and emotional tone during an interview. These
insights are valuable for providing constructive feedback that goes beyond content eval-
uation. Real-Time Feedback Systems are integrated into platforms like SmartHire to

SmartHire: AI Mock Interview Platform 13

 R.C.P.I.T.,Shirpur

deliver immediate suggestions on response structure, keyword usage, filler words, and
body language (if video is supported). This immediate loop of feedback helps users it-
erate and improve faster. Data Privacy and Access Control mechanisms are vital when
dealing with personal data like resumes, video recordings, and test results. Modern AI
platforms implement encryption protocols and user-specific access rights to ensure that
data is secure and accessible only to authorized users. Cloud Storage and Deployment
enhance scalability and availability, enabling users to access the platform from anywhere.
It also facilitates centralized management of performance reports and interview history.
User-Centered Interface Design, typically implemented using technologies like React.js,
ensures ease of use, especially for non-technical users. Integration with authentication
tools like Google Sign-In or MetaMask (in Web3-compatible systems) makes access both
secure and seamless. In summary, traditional interview preparation methods lack scalabil-
ity, personalization, and real-time feedback, which limits their effectiveness. AI-powered
platforms like SmartHire overcome these limitations through automated, intelligent, and
scalable solutions. By utilizing AI technologies such as NLP, ML, and speech analysis,
these platforms empower users with personalized insights, improve interview readiness,
and democratize access to high-quality career preparation tools.

SmartHire: AI Mock Interview Platform 14

Chapter 3

Purposed System

The proposed AI Mock Interview Platform is a comprehensive and interactive system

designed to simulate real-world interview experiences using advanced artificial intelli-
gence. It aims to improve user preparedness, confidence, and performance by offering
personalized and data-driven feedback. The system is structured into several key stages,
each contributing to a seamless and intelligent interview preparation workflow.The pro-
cess begins with the User Interface, which is accessible through both web and mobile
applications. The interface is designed to be intuitive, clean, and user-centric, allowing
easy navigation for users with varying levels of technical proficiency. This ensures that
the platform remains accessible to a broad audience, including students, job seekers, and
professionals seeking to upskill.Once the platform is accessed, users are guided through a
User Registration and Profile Setup phase. During this stage, users input basic personal
and professional details such as name, email, educational background, work experience,
skills, and career interests. This profile data is essential for tailoring the interview ques-
tions to the user’s unique background and goals. It also enables the system to track
progress over time and provide more targeted feedback in future sessions.
As shown in Fig3.1 which is a block diagram of purposed system,After setting up
their profiles, users are prompted to select the interview type, which includes choosing a
job role (e.g., software developer, data analyst, marketing executive) and the desired dif-
ficulty level (e.g., beginner, intermediate, advanced). This selection allows the system to
contextualize the interview questions and structure the session to match industry-specific
expectations and skill levels.Next, the system utilizes AI models to generate a set of AI-
Generated Interview Questions. These questions are dynamically created using a combi-

15
 R.C.P.I.T.,Shirpur

Figure 3.1: Block diagram of the proposed system

nation of Natural Language Processing (NLP) and Machine Learning (ML) techniques.
The system ensures variability and realism by incorporating behavioral, situational, and
technical questions commonly asked in actual interviews. The questions may adapt in
real-time based on the user’s previous responses, mimicking a natural conversation with
a human interviewer.Finally, the platform delivers Feedback and Scoring, providing users
with a detailed performance report. This includes scores on clarity, content relevance,
confidence, speaking pace, filler words, and more. It also offers actionable suggestions for
improvement, helping users identify their strengths and work on weaker areas. The feed-
back is designed to be constructive, encouraging, and personalized, making each session
a valuable learning experience.[6]

3.1 Working
The flowchart represents a streamlined process of an AI-based mock interview platform
designed to enhance interview preparation through intelligent automation. The system
starts with the user providing profile details and selecting preferences such as job role
and difficulty level. Using this input, the AI dynamically generates tailored interview
questions suited to the user’s goals.The user then participates in a mock interview ses-
sion through text, audio, or video, simulating a real-life interview environment. The AI
captures and processes the responses, analyzing them across several dimensions including
content quality, communication style, sentiment, and relevance. If video input is used,
optional non-verbal cue analysis is performed using computer vision.
As shown in fig 3.2, Once the analysis is complete, the system compiles a feedback

SmartHire: AI Mock Interview Platform 16

 R.C.P.I.T.,Shirpur

Figure 3.2: Workflow

report powered by AI-driven synthesis. This feedback includes actionable suggestions,

specific examples for improvement, and an optional performance score. The user can use
this detailed feedback to reflect, practice again, and gradually improve their interview
skills through repeated iterations.The working of the SmirtHire AI Interview Platform
involves several key steps that ensure the proper working of System:

1 User Interface (Web/App):The platform is accessible through both web and mobile
applications, ensuring flexibility and ease of access across devices. The interface
is designed to be intuitive and user-friendly, allowing users to navigate the system
effortlessly. It includes clear menus, prompts, and progress indicators that guide
users through each step of the interview preparation process. This accessibility
makes the platform suitable for a wide range of users, from students and job seekers
to professionals looking to practice and refine their skills.

2 User Registration Profile Setup:Upon launching the platform, users are required to
register and set up a personal profile. During registration, users provide essential de-
tails such as name, email, educational qualifications, professional experience, career
objectives, and preferred industry. This information helps the platform tailor inter-
view questions to match the user’s background and aspirations. A well-structured

SmartHire: AI Mock Interview Platform 17

 R.C.P.I.T.,Shirpur

profile enables more relevant question generation and more personalized feedback.

3 Select Interview Type, Job Role, and Difficulty: After completing their profile,
users choose the interview type they wish to simulate. Options may include HR
interviews, technical interviews, or behavioral interviews. Additionally, users select
the specific job role (e.g., software developer, marketing analyst, data scientist) and
a desired difficulty level—ranging from beginner to advanced. These selections en-
sure that the platform generates contextually appropriate and challenging questions
aligned with the target role and experience level.

4 AI-Generated Interview Questions:Once preferences are set, the system uses pre-
defined question sets and AI algorithms to generate dynamic interview questions
tailored to the selected role and difficulty. These questions may cover a range of
topics including domain knowledge, problem-solving, situational judgment, and soft
skills. The AI ensures variability and realism in question delivery, simulating the
unpredictability of real interviews.

5 Gemini AI API Processes Input: As the user responds—either through typed text or
spoken input—the Gemini AI API processes the responses using advanced Natural
Language Processing (NLP) techniques. The system analyzes linguistic patterns,
grammar, vocabulary, coherence, and the overall relevance of the answers. If audio
or video input is used, it may also assess tone, fluency, and even non-verbal cues
such as eye contact and facial expressions.

6 Feedback Scoring: After processing the user’s responses, the system provides com-
prehensive feedback that highlights strengths and areas for improvement. This
includes qualitative insights (e.g., suggestions to improve clarity or structure) and
quantitative scores based on key criteria such as accuracy, confidence, completeness,
and relevance. The feedback is designed to be constructive and actionable, empow-
ering users to reflect on their performance and make meaningful improvements with
each session.

SmartHire: AI Mock Interview Platform 18

 R.C.P.I.T.,Shirpur

3.2 Algorithm

3.2.1 Natural Language Processing (NLP) Algorithms:

Natural Language Processing (NLP) is a core component of SmartHire, enabling the

platform to understand, analyze, and evaluate human language. Algorithms such as to-
kenization, part-of-speech tagging, named entity recognition, and semantic similarity are
employed to interpret user responses. NLP helps in assessing grammar, sentence struc-
ture, tone, and relevance of answers during mock interviews. Transformer-based models
like BERT or RoBERTa are used to understand context and provide personalized feed-
back. These models are trained on large datasets and fine-tuned to understand interview
scenarios.

3.2.2 Sentiment and Emotion Analysis:

SmartHire integrates sentiment analysis algorithms to determine the emotional tone of

a candidate’s response. Techniques based on Recurrent Neural Networks (RNNs), Long
Short-Term Memory (LSTM), or attention mechanisms help classify responses into cat-
egories such as positive, neutral, or negative. Emotion recognition algorithms extend
this by identifying emotions like confidence, anxiety, or enthusiasm, which are essential
indicators in an interview setting. This analysis helps provide constructive insights and
improve user readiness.

3.2.3 Speech-to-Text (STT) and Voice Analysis:

Speech-to-text algorithms using models like Google’s Wav2Vec2 or Mozilla DeepSpeech

convert verbal responses into text for further NLP-based analysis. Additionally, voice
features such as pitch, tone, clarity, and speech rate are analyzed using signal processing
algorithms. This enables the platform to detect hesitations, filler words, or nervousness,
mimicking a real interview evaluator. These insights are used to guide users in improving
communication skills and verbal confidence.

SmartHire: AI Mock Interview Platform 19

 R.C.P.I.T.,Shirpur

3.2.4 Machine Learning-Based Scoring Systems:

SmartHire uses supervised machine learning models to evaluate and score candidate
responses. Features extracted from both text and voice inputs are fed into classifiers
(e.g., SVM, Random Forests, or Deep Neural Networks) trained to assess performance
on dimensions such as relevance, fluency, professionalism, and completeness. The model
outputs are compared against benchmark answers to generate a performance score and
feedback.[7]

3.2.5 Real-Time Feedback and Adaptive Questioning:

SmartHire leverages reinforcement learning techniques and decision trees to deliver real-
time feedback and dynamically adapt follow-up questions based on user performance.
If a candidate struggles with a certain topic, the system can rephrase or simplify the
next question. This adaptive capability enhances learning and provides a personalized
experience.

3.2.6 Data Privacy and Security Algorithms:

Given the sensitivity of user data such as resumes, voice recordings, and evaluation re-
ports, SmartHire employs standard data encryption techniques. Algorithms like AES
(Advanced Encryption Standard) ensure that stored data remains secure and confiden-
tial. Passwords and authentication tokens are protected using secure hash functions like
SHA-256. Access control mechanisms are implemented through role-based policies to
restrict unauthorized data access.

3.2.7 Recommendation System:

A collaborative and content-based filtering hybrid recommendation algorithm is employed

to suggest areas of improvement, additional practice questions, and learning resources.
This system learns from user interactions, previous scores, and interests to provide tar-
geted recommendations, enhancing the overall interview preparation journey.0

SmartHire: AI Mock Interview Platform 20

 R.C.P.I.T.,Shirpur

3.3 Software and Hardware Requirements

3.3.1 Hardware Requirements

Hardware Component Specification
Processor 1.5GHz or more
RAM 4GB Minimum
HDD Minimum 900MB free space
Operating System Windows, MacOS

Table 3.1: Hardware Requirements

Table 3.1 outlines the minimum hardware requirements necessary to ensure the proper
functioning of the software or system. The processor must have a speed of at least 1.5 GHz
to handle the software’s operations and computations effectively. A minimum of 4 GB
of RAM is required to ensure smooth performance, especially when managing multiple
tasks or running resource-intensive processes. Additionally, at least 900 MB of free hard
disk space is necessary to install the application and store its essential files, including any
temporary data it may generate during use. The software is designed to be compatible
with both Windows and macOS operating systems, providing flexibility and accessibility
for a wide range of users. These specifications serve as a baseline to guarantee optimal
performance and reliability.

3.3.2 Software Requirements

Software Component Specification
Programming Languages HTML, CSS,JavaScript
Libraries ReactJs,Gemini
Database postgress SQL, Supabase
Backend Framework typescript

Table 3.2: Software Requirements

Table 3.2 presents the software requirements necessary for the development and op-
eration of the AI Mock Interview Platform. The application utilizes HTML, CSS, and

SmartHire: AI Mock Interview Platform 21

 R.C.P.I.T.,Shirpur

JavaScript to design dynamic and responsive web interfaces that ensure a smooth user
experience. ReactJS serves as the primary library for building modular and efficient fron-
tend components. The backend logic is implemented using TypeScript, which enhances
code reliability and maintainability through its strong typing features. For intelligent
interview evaluation and response generation, the system integrates Gemini AI APIs, en-
abling advanced natural language processing capabilities. The data storage and real-time
functionalities are managed through PostgreSQL, a powerful relational database, along
with Supabase, which simplifies backend operations by offering authentication, API ser-
vices, and live data synchronization. These software components collectively contribute
to a secure, scalable, and intelligent platform that enhances interview preparation through
personalized AI-driven experiences.

SmartHire: AI Mock Interview Platform 22

Chapter 4

Methodology

The system architecture illustrated in the diagram represents the functioning of an AI-
powered mock interview platform that facilitates the secure and intelligent processing
of interview sessions. The workflow is initiated when a user accesses the platform via
a web or mobile application and inputs their profile details and preferences, including
the type of interview, desired job role, and difficulty level. This data is used by the
system’s AI-driven engine—powered by Gemini AI—to dynamically generate tailored
interview questions that are both relevant and challenging based on the user’s goals.
Once the interview begins, the user interacts with the platform through text, audio, or
video, depending on their preference. Their responses are captured and sent to the core
processing module, where multiple AI algorithms analyze the input. These include Nat-
ural Language Processing (NLP) for understanding verbal content, sentiment analysis
to assess tone and attitude, keyword matching for technical accuracy, and computer vi-
sion (if video input is used) for analyzing non-verbal cues such as facial expressions and
body language. After processing the data, the system synthesizes the insights and gener-
ates personalized feedback. This feedback includes specific suggestions for improvement,
highlights of strengths, and an optional performance score to benchmark the user’s readi-
ness. The user then receives this feedback, which allows them to refine their performance
through repeated practice. The architecture is modular and interactive, ensuring con-
tinuous communication between user inputs, AI modules, and feedback delivery. Each
transition—from question generation to analysis and scoring—is designed to enhance the
realism of interview simulations and maximize user learning outcomes. This system thus
operates not only as a mock interview tool but as a complete feedback loop that en-

23
 R.C.P.I.T.,Shirpur

4.1 System Design

Figure 4.1: System Architecture

ables users to iteratively improve their skills and confidence in a structured, secure, and
intelligent environment.
React.js :
React.js is used as the primary frontend technology, providing a dynamic and respon-
sive interface for users to interact with the AI-based mock interview platform. React’s
component-based architecture enables efficient updates to the user interface, making the
application highly responsive, especially when handling real-time processes such as AI-
generated questions, response analysis, and feedback updates. This is particularly crucial
for a system like this, where users expect immediate feedback and seamless interaction
during mock interviews. One of the key advantages of React.js is its flexibility in integrat-
ing with AI services and APIs, such as Google’s Gemini API. These integrations enable
the frontend to communicate directly with the backend services, allowing users to submit
responses (text, audio, or video), receive dynamically generated interview questions, and

SmartHire: AI Mock Interview Platform 24

 R.C.P.I.T.,Shirpur

view detailed feedback—all within an intuitive interface. This structure ensures users
can practice interviews efficiently without relying on complex or static systems. React’s
powerful state management capabilities, facilitated by tools like Redux and Context API,
allow efficient handling of global states across various components of the application. For
example, when a user logs in or selects a specific interview type, React can manage this
state and ensure that the interface reflects the correct session and context. Similarly,
when responses are submitted or scores are generated, React automatically updates the
relevant UI sections to reflect real-time changes, enhancing the user experience. Vari-
ous components of the application—such as navigation menus, interview question panels,
input forms, feedback displays, and scoring modules—are managed seamlessly through
React, ensuring consistency across the platform. React’s scalability also makes it an
ideal choice for this mock interview platform, allowing future features such as personal-
ized learning paths, multi-language support, or integrations with other AI models to be
added easily without disrupting existing functionality.
JavaScript:
In this application, JavaScript is used to create the web interface, including both user
and admin modules, and to manage their behavior on the internet. As one of the core
technologies of the web, JavaScript plays a vital role in delivering a dynamic, interac-
tive, and seamless experience for users preparing for interviews through the platform.
JavaScript is a key component in the development of the AI Mock Interview Platform,
enhancing functionality and user engagement. Being a versatile language that runs in the
browser, it is ideally suited for building interactive web applications where responsive-
ness and real-time user feedback are crucial. In this project, JavaScript enables dynamic
content updates without requiring full page reloads, improving the overall fluidity of the
platform. A primary use of JavaScript in this system is in client-side validation and in-
terface interactivity. For instance, during user registration and profile setup, JavaScript
validates form inputs to ensure data correctness before submission, thus providing im-
mediate feedback and improving the user experience. Additionally, JavaScript is used
to dynamically display AI-generated interview questions and capture real-time user re-
sponses via forms, audio, or video inputs. JavaScript also manages the integration with
AI services, such as communicating with the Gemini AI API. Through asynchronous calls
and event handling, JavaScript fetches questions, submits answers, and retrieves person-

SmartHire: AI Mock Interview Platform 25

 R.C.P.I.T.,Shirpur

alized feedback and scores—without interrupting the user flow. These interactions are
crucial to maintaining a responsive and real-time environment, simulating an authentic
interview experience. Furthermore, JavaScript contributes to the security and integrity
of the platform. It is used to ensure that sensitive data, such as user inputs and session
tokens, are securely handled within the browser. JavaScript can also facilitate secure
communication protocols using HTTPS to protect data transmitted between the client
and the server. Overall, JavaScript is instrumental in bringing the AI Mock Interview
Platform to life—enabling real-time interactivity, managing user actions, and maintaining
secure, efficient communication with backend services and AI modules.
PostgreSQL:
PostgreSQL, an advanced open-source relational database system, plays a crucial role in
the development and operation of the AI Mock Interview Platform. Known for its ro-
bustness, scalability, and support for complex queries, PostgreSQL is utilized to manage
all structured data related to user accounts, interview sessions, performance feedback,
and system configurations. One of the primary uses of PostgreSQL in this platform
is to handle user data management. When a new user registers and sets up a profile,
their information—such as name, email, selected job roles, and past interview history—is
securely stored in the database. PostgreSQL’s support for relational modeling allows
the platform to efficiently organize this data across multiple tables and ensure consis-
tency through foreign key relationships. Another essential application of PostgreSQL is
in interview session tracking and analytics. As users engage with mock interviews, their
responses, timestamps, selected question sets, and AI-generated feedback are recorded.
PostgreSQL facilitates real-time storage and retrieval of this data, enabling features such
as session history, performance trend visualization, and personalized question recom-
mendations. Complex SQL queries and indexing strategies make it possible to quickly
analyze user progress and generate insights without delay. PostgreSQL also supports the
feedback and scoring modules. The database stores evaluation metrics such as response
relevance, clarity, and language use, which are calculated by the AI engine and linked to
each user session. This data is used to generate comprehensive reports that help users
identify strengths and areas for improvement. The system’s ability to retrieve this data
dynamically ensures a smooth user experience when accessing past reports or comparing
performance over time. In addition, PostgreSQL contributes to system integrity and secu-

SmartHire: AI Mock Interview Platform 26

 R.C.P.I.T.,Shirpur

rity. With built-in roles, privileges, and authentication mechanisms, the platform enforces
secure access control, ensuring that sensitive user data is protected. PostgreSQL’s ACID
compliance also guarantees reliable transactions, preserving the consistency of interview
records and user interactions even in the event of system failures. Overall, PostgreSQL
serves as the backbone of the AI Mock Interview Platform’s data infrastructure, enabling
efficient data handling, reliable performance, and a scalable foundation to support future
enhancements such as multi-language support, advanced analytics, or integration with
third-party career services.[8]
TypeScript:
TypeScript, a statically typed superset of JavaScript, is a foundational technology used in
the development of this AI Mock Interview Platform. It enhances JavaScript by adding
strong typing, interfaces, and modern object-oriented programming features, allowing
developers to write more robust, scalable, and maintainable code. TypeScript ensures
that the codebase is less prone to runtime errors by enabling compile-time type checking,
which is especially beneficial in large-scale applications with multiple interacting modules,
such as this platform.
When building features like dynamic question generation, AI evaluation workflows,
and user session handling, TypeScript provides clear structure and improved developer
tooling, including auto-completion, refactoring support, and real-time error checking. For
example, each module—whether handling user authentication, interview configuration, or
API communication with Gemini AI—is strongly typed, allowing seamless coordination
between the frontend (built in React.js) and backend services.
TypeScript is also integral in managing asynchronous operations like API calls to the
Gemini AI engine or fetching user data from the PostgreSQL database. Its support for
modern JavaScript features such as async/await, generics, and modules streamlines the
development of responsive and performant systems that react in real-time to user input,
such as generating instant interview feedback or scoring reports.
By replacing loosely-typed JavaScript logic with TypeScript, the platform reduces the
chances of bugs caused by undefined behaviors, especially in mission-critical operations
like scoring, session persistence, and role-based access control. It also makes the applica-
tion easier to scale and extend—whether by adding new interview categories, integrating
external services, or expanding analytics modules.

SmartHire: AI Mock Interview Platform 27

 R.C.P.I.T.,Shirpur

In summary, TypeScript provides the structural integrity, clarity, and efficiency needed
to develop a complex and feature-rich AI Mock Interview Platform. Its integration along-
side other key technologies—like React.js for user interaction, JavaScript for web func-
tionality, and PostgreSQL for data management—ensures that the system is reliable,
future-proof, and ready for continuous improvement.

SmartHire: AI Mock Interview Platform 28

Chapter 5

Implementation Details

5.1 Login page Interface

Figure 5.1: Login Interface

The login interface of SmartHire AI is designed with a modern and user-centric layout
that offers both functionality and ease of use. The page features a dark-themed UI with a
minimalistic and professional appearance, aligning with contemporary design standards
for tech applications.The login form includes:

1. Email Input Field:

Allows users to enter their registered email addresses. Placeholder text provides
guidance (e.g., [email protected]).

29
 R.C.P.I.T.,Shirpur

2. Password Field:
Users can securely input their passwords. A visibility toggle icon allows users to
view or hide their password input.

3. Forgot Password Link:

Positioned next to the password field for quick access to the password recovery
process.

4. Standard Sign In:

Clicking the purple “Sign in” button submits the credentials for traditional login.

5. Google Authentication:
A white button labeled “Sign in with Google” provides a secure, OAuth-based login
via Google accounts, simplifying access for users who prefer federated login.

5.2 Dashboard

The SmartHire AI Dashboard serves as the central hub for users after success-
fully logging into the platform. It provides a sleek and interactive user experience
designed to help candidates navigate through various technical interview options
tailored to different developer roles. The design emphasizes clarity, accessibility,
and user empowerment.

Figure 5.2: Dashboard

Key Features:

• Branding and Header.

SmartHire: AI Mock Interview Platform 30

 R.C.P.I.T.,Shirpur

• Main Heading and Subtext.

• Search Functionality.

• Interview Role Cards:

– Frontend Developer

– Backend Developer

– Full Stack Developer

– Java Developer

– Python Developer

– JavaScript Developer

The use of distinct role-based colors, clean typography, and intuitive button place-
ment makes the interface both visually appealing and highly usable. The dark
theme combined with vibrant accent colors helps reduce eye strain and improves
focus, making the dashboard comfortable for extended use.

5.3 Interview Instructions

The Interview Instructions screen in SmartHire AI provides users with comprehen-

sive guidelines before beginning their AI-driven technical interview. This section
ensures candidates are well-informed about the process, expectations, and plat-
form constraints to create a professional and controlled environment that simulates
real-world interviews.[9]

5.3.1 Interview Process

• The interview comprises multiple questions specific to the user’s chosen role
and level.

• Users are given 20 minutes to complete the interview.

• Each answer must be thorough, showcasing both knowledge and problem-

solving ability.

• Upon completion, the platform generates a detailed performance analysis.

SmartHire: AI Mock Interview Platform 31

 R.C.P.I.T.,Shirpur

5.3.2 Recording Your Answers

• SmartHire AI supports both text and speech input:

– Candidates may type answers or use speech recognition.

– A ”Start Recording” button initiates speech-to-text functionality.

– Recording begins automatically after the system reads each question.

– Users must manually stop recording once the answer is completed

5.3.3 Important Rules

• To maintain integrity and focus, the platform enforces strict behavioral rules:

– Tab switching is restricted to 3 times. Exceeding this limit results in

automatic submission of the interview..

– Copy-paste is disabled to encourage original answers.

– Screenshots are prohibited, also leading to automatic submission if at-

tempted.

– The user’s camera remains active to simulate a monitored interview set-

ting.

5.3.4 Coding Questions

• A built-in code editor is provided.

• Users can write and test code directly within the platform.

• Multiple programming languages are supported, allowing flexibility for candi-

dates.

5.3.5 Exiting the Interview

• The interview can be ended at any time via the “Abort Interview” button.

• However, aborting will discard all answers, and no performance results will be
saved.

SmartHire: AI Mock Interview Platform 32

 R.C.P.I.T.,Shirpur

Figure 5.3: Instruction

5.4 Experience Level Selection

After selecting a specific job role, such as Backend Developer, the user is directed
to the Experience Level Selection screen. This interface allows candidates to choose
their expertise level, enabling a more tailored and relevant interview experience.

key Features:

• Role-Specific Context:The title of the chosen job role (e.g., Backend Developer)
is prominently displayed, confirming the user’s current track.

• Experience Levels:Users are offered three distinct tiers, each corresponding to

a different depth of technical knowledge:

– Beginner:Designed for entry-level applicants, this option includes founda-

tional questions that assess basic understanding and essential skills.

– Intermediate:Intended for mid-level professionals, this level introduces more

complex challenges and conceptual questions.

– Advanced:Geared towards experienced or senior developers, this track fo-

cuses on deep technical problem-solving and system-level thinking.

SmartHire: AI Mock Interview Platform 33

 R.C.P.I.T.,Shirpur

• Intuitive Design:Each experience level is represented with a unique gradient-

colored button and clear labels, enhancing usability and visual appeal.

• Navigation: A “Back to Job Roles” link is provided for users who wish to
change their selected job role before proceeding.

Figure 5.4: Different Levels Of Interviews

5.5 Permissions Check – SmartHire AI

Before beginning the interview, SmartHire AI performs a permissions check to en-

sure a secure and functional environment. This screen verifies the necessary hard-
ware access, including your camera and microphone.

5.5.1 Key Features

• Permission Request Popup: A browser-native dialog box appears, asking

for microphone access with options:

– Allow while visiting the site

– Allow this time

– Never Allow

• System Message: ”Checking camera and microphone permissions...”This

confirms that both devices must be enabled to proceed with the interview.

SmartHire: AI Mock Interview Platform 34

 R.C.P.I.T.,Shirpur

• Purpose of Permissions: View detailed metadata for each file, including:

– Enable speech-to-text input for answering questions.

– Activate the camera for simulating a real interview environment and en-
suring candidate presence.

• User Guidance: A brief note clarifies that permissions are required for the
interview process, helping users understand the necessity of granting access.

This screen ensures that all technical requirements are met before starting the
assessment, maintaining both the integrity and functionality of the AI-powered
interview experience.

Figure 5.5: Permissions For Camera,Audio etc.

5.6 Interview Interface

This is the main interview screen where the user actively responds to questions. The
platform is currently conducting a Beginner-level interview for a Backend Developer
role.

5.6.1 Key Features Displayed

• Time Remaining: A countdown timer (top left) shows 19:09 minutes re-
maining out of the total allotted time.

• Interview Level Tag: A green badge marks the level as ”Beginner Level”.

SmartHire: AI Mock Interview Platform 35

 R.C.P.I.T.,Shirpur

• Screen Recording Status:A purple banner indicates that screen recording

is active, ensuring integrity of the assessment.

• Live Camera Feed: On the right side, a live video of the candidate is shown
with a green ”Live” tag, used to monitor presence during the interview.

• Current Question (1/10):The platform is currently on question 1 of 10:“Tell

me about yourself and your experience as a Backend Developer.”

• Answer Input Section:

– Users can record their answer using the ”Start Recording” button.

– Alternatively, users can type their response in the input box labeled “Type
or speak your answer here...”

5.6.2 Purpose

This interface simulates a real-time AI-driven mock interview, allowing candidates

to practice:

• Speaking confidently.

• Structuring technical and personal answers.

• Managing time under a monitored setting.

Figure 5.6: Inteview Inteface

SmartHire: AI Mock Interview Platform 36

 R.C.P.I.T.,Shirpur

5.7 Feedback After Interview

The performance analysis report for the Beginner Backend Developer interview
reveals a comprehensive evaluation of the candidate’s overall performance, scored
at 6 out of 10. This score indicates a promising start for a beginner, with room
for growth in technical areas. The candidate successfully answered 2 out of 5
questions, which suggests a foundational understanding but highlights the need for
deeper knowledge in backend development concepts.

In the breakdown of specific competencies, the candidate scored 4/10 in technical

question handling, indicating a need to strengthen their grasp of backend tech-
nologies and related programming principles. On the other hand, they earned a
5/10 for clarity and structure of responses, showing average communication skills
that could be further refined for better expression of ideas. The quality of code
and problem-solving was also rated 5/10, suggesting some understanding of coding
practices, though improvement is necessary to reach higher proficiency.

One of the standout aspects of the performance was the candidate’s professionalism,
which received a perfect 10/10. This demonstrates excellent conduct, focus, and
engagement throughout the interview process, which is an important soft skill for
any role. The report also includes a recording of the interview for self-review,
allowing the candidate to reflect on their answers and delivery.

The feedback summary acknowledges this as a good performance for a beginner,

reinforcing that the candidate has a solid foundation but could benefit from deeper
study and practice. Among the strengths highlighted were professional behavior
and interview etiquette, while the main area for improvement is to deepen the un-
derstanding of beginner-level backend developer concepts. Overall, this assessment
suggests that the candidate is on the right path and, with targeted learning and
more practical exposure, can significantly enhance their backend development skills.

SmartHire: AI Mock Interview Platform 37

 R.C.P.I.T.,Shirpur

Figure 5.7: Feedback

SmartHire: AI Mock Interview Platform 38

 R.C.P.I.T.,Shirpur

5.7.1 Section-wise Performance Evaluation

• Technical Knowledge (4/10):The technical questions focused on fundamental

backend development concepts such as databases, RESTful APIs, server-side
logic, and basic coding tasks. The candidate demonstrated a basic understand-
ing of some topics but struggled with deeper or more practical implementa-
tions. The low score in this section indicates a need to strengthen conceptual
clarity and practical coding skills. To improve in this area, it’s recommended
to study backend frameworks (e.g., Express.js, Django, Flask), practice SQL
queries, and review HTTP methods and API development.

• Clarity and Structure of Responses (5/10):The ability to articulate thoughts

clearly and present well-structured responses is crucial in technical interviews.
The candidate scored an average 5/10 in this category, showing that while
the basic flow of responses was logical, there were occasional hesitations or
unclear explanations. With continued practice, mock interviews, and record-
ing self-responses, this skill can significantly improve. Using frameworks like
STAR (Situation, Task, Action, Result) can help in answering behavioral and
technical questions more effectively.

• Problem Solving and Coding Quality (5/10):This section evaluated the candi-
date’s ability to understand a problem, devise a logical solution, and implement
code efficiently. The candidate showcased some capability in breaking down
problems but faced challenges in translating the logic into efficient and bug-
free code. An important suggestion for growth here is to consistently practice
coding on platforms like LeetCode, HackerRank, or CodeChef. Emphasizing
clean code practices, error handling, and test cases can elevate the quality of
coding solutions in future interviews.[10]

• Professional Behavior (10/10):One of the most commendable aspects of the in-

terview was the candidate’s professional conduct. With a full score of 10/10, it
was evident that the candidate maintained excellent composure, attentiveness,
and sincerity throughout the interview. A neat appearance on the camera, re-
spectful tone, and consistent focus on the questions reflected maturity and
readiness to engage in professional discussions. This area is a strong founda-

SmartHire: AI Mock Interview Platform 39

 R.C.P.I.T.,Shirpur

tion that can boost overall impression, especially when paired with technical
growth.

Professional Behavior (10/10) One of the most commendable aspects of the inter-
view was the candidate’s professional conduct. With a full score of 10/10, it was
evident that the candidate maintained excellent composure, attentiveness, and sin-
cerity throughout the interview. A neat appearance on the camera, respectful tone,
and consistent focus on the questions reflected maturity and readiness to engage
in professional discussions. This area is a strong foundation that can boost overall
impression, especially when paired with technical growth.

SmartHire: AI Mock Interview Platform 40

Chapter 6

Testing

6.1 Testing Phase 1: Functional Setup and Core

Feature Validation

Testing for the SmartHire: AI Mock Interview Platform was carried out in two
structured phases to ensure the platform’s functionality, performance, and reliabil-
ity. The primary focus was to validate the core components such as the AI-driven
question engine, real-time video and screen recording functionality, and the backend
logic used for response evaluation. From the early development stages to the final
deployment, every critical feature was thoroughly tested through both manual and
automated methods to ensure a smooth and robust user experience. In the initial
testing phase, the main goal was to verify that the essential components of the
SmartHire platform were implemented correctly and interacting seamlessly. This
phase covered frontend interactions, backend processes, and AI-powered feedback
generation. The frontend, developed using React.js, was tested for responsiveness,
accessibility, and integration with the backend services. The interface was designed
to allow users to record their interviews using both video and screen sharing, re-
spond to AI-generated questions, and receive automated evaluations based on their
input. The backend, developed using Node.js and Express, handled API routing,
user authentication, and storage of interview responses. MongoDB served as the
database for managing user profiles, interview histories, and performance data.
During testing, dummy data was inserted to simulate real user interactions. The

41
 R.C.P.I.T.,Shirpur

AI module, which includes NLP-based evaluation and scoring of user responses, was
tested to ensure it analyzed textual inputs accurately based on relevance, tone, and
technical correctness. Furthermore, the screen recording feature was validated us-
ing browser-based APIs to ensure consistent recording across different systems and
browsers. The integration with third-party tools such as WebRTC for live video
capture and cloud storage services for saving recordings was verified to ensure data
was securely captured, stored, and accessible for future playback. Edge cases were
also considered during this phase. For instance, scenarios like internet disconnec-
tion during interviews, incomplete video uploads, and unsupported file formats were
simulated to test how the platform responded under failure conditions. UI elements
such as countdown timers, “Start Recording” buttons, and progress indicators were
reviewed for proper functionality and usability.

6.2 Testing Phase 2: AI Evaluation and Deploy-

ment

The second phase of testing focused on refining the AI evaluation system and de-
ploying the platform on a staging environment for end-to-end validation. The aim
was to simulate real-world user activity and ensure that the system could handle
interviews from start to finish, including question delivery, answer capture, auto-
mated evaluation, and report generation.

6.2.1 AI Scoring Validation

The AI evaluation engine was trained on a dataset of mock interview responses.

Unit tests were written to validate how accurately the model scored responses
across different categories: technical knowledge, communication clarity, profession-
alism, and problem-solving ability. The scoring mechanism was calibrated against
expert-reviewed responses to ensure fairness and consistency. For example, vague
or irrelevant answers received lower scores, while clear and structured answers were
scored higher.

SmartHire: AI Mock Interview Platform 42

 R.C.P.I.T.,Shirpur

6.2.2 Feedback Report Testing

An essential component of SmartHire is the generation of personalized feedback re-

ports. Each report includes a performance score, strength and weakness highlights,
and visual indicators of progress. These reports were tested for data consistency,
layout responsiveness, and PDF export capability. Particular attention was paid to
ensuring feedback text was meaningful and actionable.

6.2.3 Integration and End-to-End Testing

End-to-end tests were executed using automated tools such as Selenium and Post-
man to simulate the complete interview flow. These tests ensured that the user
journey—from sign-up and interview initiation to AI evaluation and report down-
load—was seamless. Integration between modules like user authentication, question
rendering, audio/video capture, and AI evaluation was also thoroughly validated.

6.2.4 Performance and Load Testing

To evaluate the platform’s scalability, load testing was performed using JMeter.
Simulated concurrent users were run to ensure the platform could handle multi-
ple interviews simultaneously without performance degradation. Stress tests also
revealed memory usage patterns and helped in optimizing backend APIs.

6.2.5 Final Deployment and Bug Resolution

Upon successful completion of all testing tasks, the platform was deployed on a
test server for pilot use. Bug reports generated during this phase were documented
and fixed promptly. Issues such as audio sync errors, delayed video uploads, and
occasional AI evaluation mismatches were resolved through patch updates. The
system was finally validated for GDPR-compliance and data privacy standards to
ensure user recordings and evaluation data were securely handled.

SmartHire: AI Mock Interview Platform 43

Chapter 7

Results and Discussion

7.1 Results

The Secure Sphere platform successfully implements a decentralized and The SmartHire:
AI Mock Interview Platform successfully simulates real-time interview environments
using AI, providing automated evaluation, personalized feedback, and an intuitive
interface. The platform was tested with multiple users for technical positions such
as backend development, and the following key functionalities were verified:

7.1.1 Interview Experience

• Users were able to seamlessly participate in mock interviews via a web interface
that included real-time video recording and screen-sharing capabilities.

• The AI-driven question engine dynamically generated relevant questions based

on the selected job role and difficulty level.

• Countdown timers and UI prompts ensured that users were guided properly
throughout the interview process.

7.1.2 Automated Evaluation and Scoring

• The platform accurately scored user responses across four key categories: tech-
nical understanding, communication clarity, professionalism, and problem-
solving.

44
 R.C.P.I.T.,Shirpur

• Natural Language Processing (NLP) models processed text and speech data
to assess the relevance, coherence, and tone of user answers.

• Feedback was automatically compiled into a performance report, which in-

cluded a numerical score (e.g., 6/10), highlights of strengths, and areas for
improvement.

7.1.3 Report Generation and Download

• Users received a detailed performance report immediately after the interview,

containing feedback summaries, visual scores, and personalized suggestions.

• Reports were downloadable in PDF format and could also be stored in the
user’s dashboard for future reference and improvement tracking.

The system functioned as intended, successfully delivering AI-driven interviews,

evaluating responses in real time, and producing high-quality feedback reports.
These results demonstrate the platform’s effectiveness in preparing candidates for
real-world interviews, especially for beginner-level roles.

7.2 Discussion

7.2.1 AI-Powered Feedback and Evaluation

The use of AI and NLP algorithms allowed the SmartHire platform to provide
unbiased, consistent, and structured evaluation of user responses. Unlike manual
assessments, the AI module offered immediate insights into the user’s performance
based on linguistic and contextual analysis. While the scoring system was generally
accurate, further refinement could help address edge cases like nuanced human
expressions or vague responses.

Additionally, incorporating audio sentiment analysis and prosody recognition could

further improve feedback on tone and confidence, which are important for inter-
views.

SmartHire: AI Mock Interview Platform 45

 R.C.P.I.T.,Shirpur

7.2.2 Interview Recording and User Interface

The frontend, built with React.js, enabled users to navigate the interview pro-
cess smoothly. Features like live camera feeds, progress indicators, and automatic
recording minimized confusion and created a realistic mock interview environment.
The integration of WebRTC and MediaRecorder APIs ensured that video and screen
recordings were captured reliably across browsers.

However, user feedback indicated that additional features such as pausing/resuming

interviews or previewing recordings before submission could enhance usability. Im-
provements to the UI, including mobile optimization and accessibility support, are
areas of focus for future versions.

7.2.3 Report System and User Engagement

The automated report generation was a strong differentiator of the SmartHire plat-
form. By highlighting strengths (e.g., professionalism, communication clarity) and
areas for improvement (e.g., technical depth, structuring responses), users were able
to reflect and grow.

Nonetheless, some feedback reports were found to be overly generic, especially for
users with intermediate skills. Adding a more adaptive learning model that adjusts
feedback based on user history or past performance could improve personalization.

SmartHire: AI Mock Interview Platform 46

Chapter 8

Conclusion

The SmartHire: AI Mock Interview Platform has successfully achieved its intended
objectives, offering a reliable, intelligent, and accessible system for conducting mock
interviews. This project highlights the transformative potential of artificial intelli-
gence and natural language processing in modernizing traditional interview prepara-
tion methods. By harnessing the power of AI-driven evaluation, SmartHire provides
immediate and personalized feedback, helping users identify their strengths and ad-
dress areas that need improvement. The integration of advanced technologies such
as NLP, real-time video processing, automated scoring, and user-friendly frontend
interfaces ensures a seamless experience that mimics real-world interview scenar-
ios. The platform significantly addresses a growing demand among job seekers and
students for effective and scalable interview preparation tools. Its intelligent feed-
back system, dynamic question generation, and auto-evaluation modules reduce
dependency on human interviewers while maintaining consistency and objectivity
in assessments. Data security, user privacy, and performance reliability were also
important considerations. Through careful architectural choices and efficient re-
source handling, SmartHire maintains a secure and smooth user experience. The
successful development and deployment of this platform represent a meaningful
step toward applying AI technologies in career development, education, and HR
technology. Overall, SmartHire demonstrates how artificial intelligence can democ-
ratize access to quality interview preparation and provide an effective, scalable, and
impactful solution for job readiness in today’s competitive environment.

47
Bibliography

[1] S. Rani and M. Gupta, “AI-Based Career Guidance and Interview Preparation
Using Chatbot Systems,” Procedia Computer Science, vol. 210, pp. 158–164,
2022.

[2] S. Chakraborty and S. Bhattacharyya, “AI-based Interview Preparation Plat-

form Using NLP and Semantic Matching,” International Journal of Computer
Applications, vol. 176, no. 20, pp. 17–22, 2020.

[3] A. Singh and P. Sharma, “Design and Implementation of AI-Driven Interview

Assistant Using NLP and Deep Learning,” International Journal of Artificial
Intelligence and Applications (IJAIA), vol. 13, no. 3, pp. 12–21, 2022.

[4] R. Sharma and A. Kaur, “Natural Language Processing Based Virtual Inter-
viewer Using Machine Learning,” International Journal of Scientific Research
in Computer Science, Engineering and Information Technology, vol. 7, no. 1,
pp. 144–150, 2021.

[5] S. Mishra and N. Singh, “Intelligent Interview System Using AI and Speech
Recognition,” in Proc. Int. Conf. Computing, Communication, and Intelligent
Systems (ICCCIS), IEEE, pp. 458–463, 2021.

[6] Q. Zhang and F. Yang, “Real-Time Feedback Mechanisms in AI-Based Mock

Interview Tools,” ACM Transactions on Interactive Intelligent Systems, vol.
11, no. 3, pp. 1–20, 2021.

[7] V. Kumar and A. Thomas, “Speech Pattern Analysis for Interview Evaluation
Using NLP and Audio Features,” International Journal of Speech Technology,
vol. 24, no. 1, pp. 12–23, 2021.

48
 R.C.P.I.T.,Shirpur

[8] T. Lee and J. Wang, “Emotion Recognition in AI Interview Systems Using

Multimodal Data,” in Proc. Int. Conf. on Affective Computing and Intelligent
Interaction (ACII), IEEE, pp. 310–317, 2020.

[9] M. Das and P. Banerjee, “Automated Interview Assessment Using Deep Learn-
ing and Prosodic Feature Extraction,” Journal of Artificial Intelligence Re-
search, vol. 69, pp. 101–119, 2021.

[10] A. Rahman and S. Roy, “Development of Virtual Interview Bots with Contex-
tual NLP Capabilities,” Advances in Human-Computer Interaction, Hindawi,
vol. 2022, Article ID 9876543, pp. 1–12, 2022.

SmartHire: AI Mock Interview Platform 49