AI/ML Engineer & Full-Stack Developer
I build production-ready software focusing on multi-agent AI orchestration, LLM integration, and scalable full-stack architectures. I bypass theoretical constraints by directly engineering, deploying, and maintaining end-to-end systems.
- AI Agent Architecture: Designing and orchestrating multi-agent swarm systems (super-agent coordination, micro-agent delegation) using LLM APIs (OpenAI, Gemini, Groq).
- Full-Stack Ecosystems: Building modular, microservices-style platforms split across dedicated API, web UI, and mobile repositories.
- Client & Production Delivery: Delivered a custom billing and accounting system for a private client engagement — built and maintained under a private repository per client agreement.
- Production Deployment: Managing the lifecycle of applications from local model testing and API integration to live web and mobile environments.
- Languages: Python, TypeScript, JavaScript, C++, SQL, HTML/CSS
- AI/ML: TensorFlow, PyTorch, scikit-learn, Keras, Hugging Face, OpenCV, LangChain, Multi-Agent Swarm, RAG
- LLM APIs: OpenAI API, Gemini API, Groq API
- Backend & APIs: FastAPI, Flask, Django, REST APIs, WebSockets, AsyncIO, JWT Auth
- Frontend & Mobile: React, TypeScript, React Native (Expo), Streamlit, Gradio, PyQt6, Gatsby (JAMstack)
- Databases: PostgreSQL, MySQL, SQLite, MongoDB, Supabase, Redis
- Infrastructure & Tools: Docker, GitHub Actions, Git, Linux, Jupyter, Selenium
- Shik-Shak EdTech Ecosystem: A cross-platform application supporting JEE and NEET. Features AI-powered question generation via Groq LLM and real-time battle rooms. Architected across independent API, web, and mobile repositories.
- Swarm AI Agent (Travily): A multi-agent AI system accessible through mobile app, web, and discord-bot interfaces, all backed by one shared API. A super-agent delegates tasks across specialized micro-agents, with Tavily integration for real-time search.
- Client Billing & Accounting System: A custom billing and invoicing platform delivered for a private client engagement. Core implementation is private per client agreement; a public placeholder repo documents the project.
- Medicine Scanner API & UI: A computer vision pipeline utilizing OpenCV for OCR and image processing, designed with a clean separation of concerns between backend APIs and the client interface.