Harsh Yadav

AI/ML Engineer · LLMs · RAG · MLOps

---

I build production-grade AI systems — RAG pipelines, LLM agents, and async backends that work at scale.

Currently shipping Salesforce + cloud integrations at Codm Software while building AI infrastructure on the side.

focus = {
    "building":   ["RAG Pipelines", "LLM Agents", "Multi-Agent Systems"],
    "stack":      ["Python", "FastAPI", "LangGraph", "PostgreSQL", "Docker"],
    "interests":  ["LLM Optimization", "Vector Search", "MLOps"],
    "open_to":    "AI/ML Engineer roles"
}

---

🔨 Featured Projects

🧠 Enterprise Document Intelligence Platform

RAG-based document Q&A system with async processing and semantic search

• Multi-agent architecture with LangGraph for long-context workflows
• Async task pipeline via Celery + Redis — keeps API non-blocking under load
• Vector search with ChromaDB + pgvector for sub-second retrieval
• Fully containerized with Docker, CI/CD via GitHub Actions
• Python FastAPI Celery Redis LangGraph ChromaDB Docker

---

🔍 Reddit Persona Analytics Pipeline

End-to-end ETL pipeline that profiles Reddit users using NLP

• Extracts user data via Reddit API, stores in relational DB
• Text preprocessing and behavioral pattern analysis with Pandas + NLTK
• SQL-driven reporting on user persona traits
• Python Pandas Scikit-learn NLTK PostgreSQL ETL

---

🛠 Tech Stack

AI / ML LLMs RAG LangGraph Prompt Engineering pgvector ChromaDB

Backend Python FastAPI Celery Redis PostgreSQL SQLAlchemy REST APIs

DevOps & Cloud Docker GitHub Actions AWS S3 CI/CD Render

---

📜 Certifications

• ☁️ AWS Academy Cloud Architecting — Jan 2025
• 🤖 Artificial Intelligence Foundations — SkillUp, Nov 2023
• 📊 Foundations of Data Science — Google, Oct 2023

---

B.Tech in Computer Science (AI & ML) · Krishna Engineering College · 7.75 CGPA