█████╗ ███╗   ███╗███████╗██╗   ██╗ █████╗
██╔══██╗████╗ ████║██╔════╝╚██╗ ██╔╝██╔══██╗
███████║██╔████╔██║█████╗   ╚████╔╝ ███████║
██╔══██║██║╚██╔╝██║██╔══╝    ╚██╔╝  ██╔══██║
██║  ██║██║ ╚═╝ ██║███████╗   ██║   ██║  ██║
╚═╝  ╚═╝╚═╝     ╚═╝╚══════╝   ╚═╝   ╚═╝  ╚═╝

• Building a Next.js AI Literacy Portal at TCET Centre of Excellence and internal automation tools in Python (May 2026)
• AI Systems Evaluation Intern at AirDawg Labs, creating and refining tasks used to train large language models on Snorkel, deployed for OpenAI and Anthropic (Jun 2026)
• Maintaining Mustel, a Python CLI + MCP server with 2,100+ PyPI downloads
• Founded The Free University with 90,000+ lifetime visits and 1,000+ LinkedIn followers

Static analysis pre-processing layer for AI coding agents (Cursor, Claude Code, Windsurf)

• Orchestrates 4 engines: Ruff, Bandit, pip-audit, 21 custom YAML patterns
• Outputs a normalized JSON report, eliminates AI-hallucinated bug reports
• Cuts LLM output token usage by ~45% per review cycle
• Published on PyPI with MCP server support

mustel run          # full analysis
mustel run --mcp    # MCP server mode for agent integration

FastAPI service for bulk PDF and DOCX footer stamping

• Handles up to 10 concurrent uploads
• Per-page and per-section footer editing
• ZIP or individual file download output
• Built for college students submitting lab files

Flask image compressor and PDF merger

• In-memory BytesIO file handling, no disk writes
• UUID-based download tokens for safe concurrent requests
• Image compression, resize by percentage, multi-image PDF merge
• Deployed on Render

Monte Carlo stock price simulation engine. Stateless REST API.

• Simulates future price paths using Geometric Brownian Motion (GBM)
• FastAPI backend with vectorized NumPy compute, 1,000 paths in milliseconds
• Rate limiting + Docker Compose full stack (backend, frontend, worker)
• Free and zero-auth; drop the endpoint directly into any dashboard or app

Rolling window statistics for NumPy arrays. No pandas needed.

• mean, std, sum, min, max, zscore, normalize, apply with one consistent API
• Uses numpy.lib.stride_tricks for zero-copy memory views, no data duplication
• Handles missing data via min_periods, raises clear errors on bad input
• Published on PyPI with CI, full docs site, CHANGELOG, CONTRIBUTING

Free course aggregator built and SEO-optimized from scratch

• 90,000+ lifetime visits
• 1,000+ LinkedIn followers
• Organically grown, zero paid promotion

AI Systems Evaluation Intern · AirDawg Labs (Jun 2026 – Present) Creating and refining RLHF tasks to evaluate and train large language models. Working end-to-end on the task lifecycle from ideation to final submission on Snorkel, used to improve models for OpenAI and Anthropic. Ubuntu / WSL, Docker.

Software Intern · TCET Centre of Excellence (May 2026 – Present) Contributing to a Next.js web portal for AI literacy training. Building Python-based internal automation tools.

Research Analyst → Associate Data Analyst · CoreLayer Labs (Review.AI) (Oct 2024 – Feb 2025) Built an LLM-powered product-page scraper cutting manual review time by 75%. Built a real-time Streamlit dashboard integrated with Google Sheets for team workload tracking. Researched and documented 50+ AI tools.

Languages  ·  Python, JavaScript, SQL

Backend  ·  FastAPI, Flask, Next.js, REST APIs, LLM APIs (OpenAI, Anthropic)

Tooling  ·  Git, PyPI, Docker, Ubuntu / WSL, MCP Server Integration, Ruff, Bandit

Data  ·  Pandas, NumPy, Streamlit

Deploy  ·  Render, Vercel, Streamlit Cloud

• Problem Setter · Shastra Coding Club, TCET. Designs and validates test cases for T&P coding assessments
• Mumbai Hacks 2025. Built LifeLink (AI organ donation platform), qualified for Round 2 offline of India's largest agentic AI hackathon (HealthTech track)
• CGPA 9.21 / 10 · B.E. Information Technology, Mumbai University