Pranav Kumar

I build production-grade agentic AI systems and ML models with 2.8 years of experience across fintech and enterprise use cases. I specialize in deploying full-stack AI solutions with LangGraph, FastAPI, and AWS ECS Fargate, and I’ve delivered measurable impact—improving data throughput by 30% and reducing manual effort by 50% with automated multi-agent RAG frameworks.

In my recent work, I developed LangGraph-based multi-agent chatbots and RAG applications, engineered PySpark pipelines on Databricks for 10M+ record datasets, and shipped predictive credit-risk and propensity models with 20%+ accuracy improvement over baseline. I’m comfortable with async workflows and cross-timezone collaboration, and I focus on reliable MLOps—CI/CD with GitHub Actions, observability, drift detection, and Human-in-the-Loop controls.