🌐 Portfolio • LinkedIn • GitHub
I'm an AI/ML Engineer focused on building and deploying practical AI systems.
- 🤖 Agentic AI, LLMs & RAG
- 🧠 Machine Learning & NLP
- ⚡ FastAPI, Python & SQL
- 🔗 LangChain & LangGraph
- 🛠️ AI APIs, automation & backend systems
- ☁️ Docker, AWS & deployment
- 📊 LLM evaluation, reliability & observability
AI-assisted candidate intelligence for recruiters: upload resumes, extract structured candidate profiles, score them against a job's requirements across five weighted dimensions (skills, semantic relevance, experience, education, projects), and get a recruiter-readable explanation of why a candidate is fit.
Medical document & consultation intelligence using Faster-Whisper, Groq, Pydantic, FastAPI and Docker.
Agentic weather assistant built with LangGraph, tools, retrieval and LangSmith.
Expertise Fraud Detection System
proof-of-concept for detecting potential expertise fraud in candidate profiles using machine learning, reinforcement learning, and workflow orchestration. This system analyzes profile histories, screening answers, and web signals to identify suspicious claims, providing explainable decisions for hiring processes.
🛡️ AI Fraud & Risk Agent
Multi-agent fraud/risk analysis system using CrewAI, FastAPI, PostgreSQL, Redis and Celery.
📄 Enterprise RAG
Production-oriented document QA with vector search, RBAC, retrieval evaluation and grounded generation.
Languages: Python • SQL • JavaScript
AI/ML: PyTorch • Scikit-learn • Pandas • NumPy
GenAI: LangChain • LangGraph • RAG • LLMs • Agentic AI • LoRA/PEFT
Backend: FastAPI • Flask • Pydantic
Databases: PostgreSQL • MySQL • Qdrant • FAISS • pgvector
DevOps: Docker • AWS • GitHub Actions • Git
Automation: n8n • Streamlit • LangSmith
Mental Model for Designing Production-Ready Agentic AI Systems Medium
Exploring agent orchestration, tools, memory, evaluation, reliability, monitoring and governance.
Building intelligent systems that are useful, reliable and production-ready.