Thanks to visit codestin.com
Credit goes to www.ajitkumar.io

Welcome to My Galaxy

A space to host my Personal / Professional works.
Let's Explore !!....

AK

Hi, I am Ajitkumar Senthil Kumar
|

My Fields of Interests:
|

Dear
|

From the precision of :
|

To the Logic's of :
|
My journey has been anything but ordinary. I’m a passionate engineer driven by curiosity, with roots in Mechanical Engineering and a thriving ambition in the world of Data Science, Machine Learning, and Artificial Intelligence.
Actively looking for full time opportunities in Machine learning engineer, data scientist roles, I would appreciate you to explore my works! down below

University at Buffalo

Master of Science in Data Science

New York

+1-(716)-249-9240

Ajitkumar's photo
Ajitkumar

@ajitkumar

Available

Hire Me

About

A Brief

A passionate engineer driven by curiosity and a love for innovation. I explore the evolving world of AI, always seeking breakthroughs that shape the future. With a focus on emerging trends and tech's societal impact, I thrive on turning complex ideas into exciting possibilities.

Passion

Passion towards innovation and improvement, interested in exploring the world of agents in AI, ardent on building applications with seamless integrations of the latest technologies.

Career Aspirations

I'm deeply thrilled by the possibilities of Machine Learning and aspire to build models that solve real-world problems at scale to deployment—driving impact across industries. Eager to collaborate in teams where I can contribute to both experimentation and production-ready solutions.

Education

Pursuing a Master of Science in Data Science at the University at Buffalo, with a strong focus on machine learning, deep learning, and statistical modeling. Also holds a Bachelor's in Mechanical Engineering from Anna University (Sri Sai Ram Engineering College).

Coursework

Through hands-on projects and advanced coursework, I have gained expertise in building scalable ML systems, analyzing large datasets, and applying AI techniques to real-world challenges.

Current Work

I'm currently exploring and working in reinforcement learning to further expand my knowledge in deep learning through reward-based learning enabling machines to learn efficiently and adaptively.

Experience

Skills

Languages I use
Python
R
TypeScript
SQL
PostgreSQL
Libraries I use
PyTorch
HuggingFace
MLflow
Hopsworks
Scikit-learn
Ollama
SnowFlake
LangChain
LangGraph
Tools I use
React
Next.js
Astro
Tailwind CSS
FastAPI
Streamlit
Docker
AWS
Git
GitHub
Github Actions
Supabase
Vercel

Projects

Smart Loan Predictor

Built a Production complete Full-stack end-to-end ML application that predicts the approval rate of Loan , provides a automated LLM generated report according to the response....

PythonTypescriptReactTailwindcssSci-kit LearnFastAPIDockerDagshubMLflow

MIDAS Skin Cancer Detection

Developed a CNN-based image classifier alongside a colleague for identifying melanoma using the standford skin cancer dataset. Achieved 85%+ accuracy with SE,CBAM blocks and LR scheduler.

PyTorchCNNRestNetTransformersDistil-BERTHyperparameter TuningExplainable AIData Augmentation

CovDet

An Application where you upload your chest x-ray images, it detects if the sample is a COVID-19 affected or other lung disease!, useful medical application for both patient and doctors

PyTorchCNNRestNetConvNextHyperparameter TuningExplainable AIData Augmentation

BlogPosts

A Blogs, Journals and detailed Description of all my and upcoming works!, in a blog fashion with AI summarizer!

ReactTailwindCSSTypeScript

Jersey City Citibike rides Prediction

Built an end-to-end MLops pipeline using real-time Citi Bike data with preprocessing in Hopsworks, tracked via MLflow, and deployed on Streamlit,Integrated CI/CD to fetch and load Daily data and updates in model.

PythonMLflowHopsworksStreamlitGithub Actions

New York City Taxi-rides Prediciton

Built an end-to-end ML pipeline using real-time NYC data with preprocessing in Hopsworks, tracked via MLflow, and deployed on Streamlit.

PythonGithub ActionsMLflowHopsworksStreamlit

Simulation of a Prosthetic Robotic limbs using Reinforcement Learning

Trained a humanoid-V4 robot to walk and balance focusing on the stability and efficiency of the limbs, to aid the physically challenged individuals to walk and perform daily activities with the implementation of Reinforcement Learning.

PythonPyTorchReinforcement LearningOpenAI GymTensorBoard,MuJoCoSoft Actor-CriticTD3PPO

Nexomnis

An agentic AI knowledge platform that lets users upload documents, ask questions, and receive grounded answers from their preferred LLM. Nexomnis combines Retrieval-Augmented Generation, LangGraph-based orchestration, multi-model routing, and document-aware reasoning to deliver citation-backed responses across OpenAI, Gemini, Claude, and local Ollama models.

PythonLangChainLangGraphNext.jsTypeScriptTailwindCSSChromaDBOpenAIGoogle GeminiAnthropic ClaudeOllamaRAGAgentic AI

Tutorly

An AI-powered tutoring platform that uses reinforcement learning to personalize curriculum sequencing—where students upload their syllabus, an RL agent dynamically selects optimal topics to teach, an LLM delivers lessons, and adaptive quizzes continuously optimize the learning path.

PythonLangchainLanggraphNext.jsTailwindCSSTypeScriptChromaDBReinforcement LearningSACOpenAI

Contact Me

Whether you're interested in collaborating, hiring, or just want to connect —feel free to reach out. I'm always open to new opportunities in AI, machine learning, and full-stack development.