About Projects Skills GitHub Contact
Available for opportunities

Udit Panwar

Full-Stack Developer & AI Engineer

I build AI-integrated web applications — from legal document generators to medical image annotators — with a focus on the Indian market.

udit@dev:~$

Who I am

I'm a final-year student and developer based in Noida, Delhi NCR, building full-stack products that bridge AI capabilities with real-world needs. My work spans legal tech, medical tools, and fintech — all with a focus on the Indian market.

My stack is built around Python (Flask, FastAPI), React, and cloud deployment on Render, Railway, and Vercel. I love tackling hard problems — from RAG pipelines and YOLO-based image annotation to debt reconciliation algorithms and UPI payment flows.

Previously interned as a Data Science Intern at CodeCraft Infotech (Jul–Aug 2025), where I cleaned 10K+ records, built ML/NLP models hitting ~82% accuracy, and contributed to a Kaggle MAP competition. Currently open to SDE, full-stack, or AI/ML engineering roles.

5+
Production Projects
3
Cloud Platforms
7.0
CGPA — B.Tech CSE
🇮🇳
India-Focused

What I've built

A selection of production projects — all deployed, all tackling real problems.

⭐ Featured AI / Legal
NyayaAI
AI-powered Indian legal document generator with a 30-route FastAPI backend, RAG pipeline using FAISS + SentenceTransformers, Hindi/Devanagari PDF export, and a React SPA. Fully deployed on Render + Vercel.
FastAPIReactFAISSClaude APISentenceTransformersFPDF2Pydantic
⭐ Featured AI / Medical
MediMark AI
Medical image annotation platform using Flask, OpenCV, YOLO, and U-Net segmentation. Multi-user with PostgreSQL + Cloudinary storage. Fixed a critical cross-user data leakage bug before production launch.
FlaskOpenCVYOLOU-NetPostgreSQLCloudinaryRender
FinTech Full-Stack
BillSplitter
Splitwise-style shared expense manager with a greedy debt reconciliation algorithm, UPI QR code generation, Flask/SQLAlchemy backend (MySQL + SQLite fallback), and a vanilla JS SPA frontend. Deployed across Vercel + Railway.
FlaskSQLAlchemyMySQLVanilla JSUPI QRRailwayVercel
Productivity 2-hour build
Team Task Manager
Zero-dependency Flask app built under a 2-hour constraint using stdlib only (no pip). Role-based access control, deployed on Railway. Demonstrates rapid prototyping and deployment under pressure.
Flaskstdlib-onlyRBACRailway
Data Analysis Business
Cloud Kitchen Market Intelligence
Pandas-based data cleaning pipeline and business analysis on a restaurant dataset from Nashik / Delhi NCR. Surfaced actionable market insights for the cloud kitchen industry using Python data tooling.
PythonPandasMatplotlibSeaborn

Tech stack

Tools and technologies I work with regularly.

Backend
Python
Flask
FastAPI
REST APIs
SQLAlchemy
Frontend
React
JavaScript
HTML / CSS
Vanilla JS SPA
AI / ML
Anthropic API
FAISS / RAG
YOLO / OpenCV
SentenceTransformers
Scikit-learn / NLP
Data
PostgreSQL
MySQL
SQLite
Pandas
Cloud / DevOps
Render
Vercel
Railway
Cloudinary
Git / GitHub
Tooling
C++ / OOP
FPDF2 / python-docx
Pytest
PowerShell / CLI

Contributions

UP
uditpanwar261 @GitHub
Commits (est.)
5+
Repos
Day streak
Less
More

Get in touch

I'm actively looking for full-stack, SDE, or AI engineering roles. Whether it's a question, a project idea, or a job opportunity — I'm quick to respond.