Thanks to visit codestin.com
Credit goes to github.com

Skip to content

AI-powered analyzer for U.S. visa applications (J1, DS-160). Upload documents, assess risks, get recommendations. Built with Flask + OpenRouter (Qwen-23B).

fromgodd/US-Visa-Agent

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

https://us-visa-agent.onrender.com/ AI-powered analyzer for U.S. visa applications (J1, DS-160). Upload documents, assess risks, get recommendations. Built with Flask + OpenRouter (Qwen-23B).

VisaAssist — US Visa Risk & Feedback Analyzer 🇺🇸🛂

VisaAssist is a Flask-based web app designed to analyze US visa applications (especially J1, DS-160) and provide:

  • Rejection probability estimates
  • Honest assessment of strengths and weaknesses
  • Actionable advice to improve the application

It uses:

  • OpenRouter AI (Qwen 3 235B model) for deep natural language analysis
  • Custom rules engine for base risk scoring (incl. logic for countries like Uzbekistan)
  • PDF parsing for extracting content from uploaded visa documents

🔧 Features

  • Upload and process visa-related PDFs (financial docs, DS-160, etc.)
  • Smart extraction and formatting of applicant data
  • Chat-style prompt to a large language model for analysis
  • HTML-formatted AI feedback (clear, structured, and exportable)
  • Dynamic rejection probability calculator based on real-world logic

🚀 Getting Started

1. Clone this repo

git clone https://github.com/fromgodd/US-Visa-Agent.git
cd US-Visa-Agent

2. Set up your environment

Create a .env file with:

OPENROUTER_API_KEY=your_openrouter_api_key_here

Get your free API key at https://openrouter.ai/

3. Install dependencies

pip install -r requirements.txt

4. Run the app

python app.py

Visit http://localhost:5000 to access the web interface.


📈 Visa Rejection Logic

The app uses a country-based baseline risk (e.g., Uzbekistan ~ 64.41%) and applies weighted risk factors like:

  • Weak ties to home country
  • Poor financials
  • Lack of interview prep
  • Incomplete documents
  • Previous visa issues

Each factor increases the final probability (capped at 95%).


🧠 AI Model

We use the free qwen/qwen3-235b-a22b:free model from OpenRouter API for feedback generation.


🗂 Tech Stack

  • Python + Flask
  • HTML (Jinja2 templating)
  • OpenRouter AI (Qwen 3 model)
  • PyPDF2 for PDF parsing
  • RESTful JSON API communication

📜 License

MIT License. Do whatever you want, just don’t blame us if you get rejected at the embassy 💀


🙋‍♂️ Author

Made with love and skepticism by a math-loving Rust-enjoyer named Valera 🇺🇿🛠️

+ Pull requests are welcome
+ TODO - Fine tune for better results!
- Dumb visa denials are not

Screenshots

{66832B7A-7974-4074-9028-6DBAA896F38D}

{B552203C-D188-4A83-B706-BA5BFC40F3A5}

About

AI-powered analyzer for U.S. visa applications (J1, DS-160). Upload documents, assess risks, get recommendations. Built with Flask + OpenRouter (Qwen-23B).

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published