Praveen Kumar

About Praveen Kumar

I\'ve trained developers. The single question they all ask me now is the same:\"How do I stay relevant when AI can write code?\"Here\'s what I tell them — and what I believe: the developers who win in the next 5 years won\'t be the ones who resist AI. They\'ll be the ones who learn to design systems where AI operates reliably in the real world. That skill is still rare. I\'m building it, teaching it, and sharing everything I learn along the way.By day, I\'m a Senior AI Engineer at JPMorgan Chase, where I build production-grade agentic systems, LLM-powered workflows, and AI-native full-stack platforms that handle real financial complexity at scale.By night (and weekends), I break down what\'s actually working in AI engineering for the 5K+ developers I\'ve taught through workshops, courses, and the Engineering Coders community I founded in 2022.MTech in AI from IIT Jodhpur. AWS Certified Developer (2025). I\'ve seen this space from both sides — the cutting edge of enterprise AI and the classroom where developers are trying to catch up.What I focus on: LangGraph, Agentic AI, MCP (Model Context Protocol), RAG systems, LLM fine-tuning, FastAPI, and building full-stack platforms that actually ship.If you\'re a developer trying to navigate the AI shift — follow along. I post breakdowns of what I\'m building, what\'s failing, and what\'s working, straight from the production floor. Want to work together or have a technical question? Connect or DM me directly.

Experience

1. Senior Software Engineer

   摩根大通

   Dec 2025 — Present · Hyderabad, IN

   Building AI-native backend systems at one of the world\'s largest financial institutions, focusing on agentic AI and intelligent automation at scale.Key contributions:• Architected and deployed multi-agent LLM pipelines using LangGraph, reducing manual document processing time by ~70% across compliance workflows• Built RAG-based internal knowledge systems using vector databases (Pinecone/pgvector), cutting average query resolution time from hours to under 2 minutes• Designed FastAPI microservices handling 50K+ daily transactions with 99.9% uptime, integrated with JPMorgan\'s internal MCP tool ecosystem• Led adoption of Model Context Protocol (MCP) for standardizing AI tool integrations across 3 internal teams• Mentored 4 junior engineers on LLM fine-tuning, agentic design patterns, and production AI deploymentStack: Python, LangGraph, LangChain, FastAPI, Node.js, OpenAI/Claude APIs, Pinecone, PostgreSQL, AWS, Docker, Kubernetes

Education

• Indian Institute of Technology Jodhpur

  Master of Technology - MTech, Artificial Intelligence

• Rajiv Gandhi Prodyogiki Vishwavidyalaya

  Bachelor of Engineering - BE, Electrical, Electronics and Communications Engineering

  2015 — 2019