Harkirat Chahal

I’m a final-year CS undergraduate specializing in AI/ML (TIET, 2026) who ships production-grade agentic AI systems. I focus on end-to-end delivery: model training, FastAPI/FastMCP services, LangGraph agents, and async workers deployed on Railway.

I’ve built LLM-driven pipelines that combine hybrid retrieval and reranking for practical outcomes, including an AI recruitment workflow with BM25 + Gemini-embedding RRF, LightGBM ranking, and LLM match explanations. I also build AI security tooling using MCP servers for autonomous vulnerability scanning and contextual risk scoring.

Beyond core AI work, I contribute to backend and infrastructure: FastAPI + SQLAlchemy/Alembic with PostgreSQL/SQLite, caching and async processing with Celery/Redis, and containerized deployments with Docker/Kubernetes. In a small agile team, I’ve supported a Kotlin/Jetpack Compose app and refactored Go APIs to cut latency by up to 30% on high-traffic endpoints.