Open to opportunities

Narayana koushik Kancharla

@narayanakoushikkanch

Message

I build scalable multi-agent AI systems and reliable automation workflows.

United States

Message

What I'm looking for

I'm looking to build and productionize scalable AI agents and automation workflows, with opportunities to work on multi-agent systems, reliable evaluation, human-in-the-loop controls, and continuous improvement.

I've built multi-agent AI systems at Quanta Manufacturing and Lillup, using Python, LangGraph, and LangChain to automate diagnostic and learning workflows. At Quanta Manufacturing, I designed a diagnostic system that processed over 10,000 server logs and supported fault handling across 30+ concurrent systems.

At Lillup, I orchestrated AI workflows handling over 10,000 user interactions monthly and developed FastAPI endpoints for retrieval-augmented generation that improved personalized content relevance by 25%. I also built automated evaluation frameworks that reduced manual review time by 60% while improving response consistency.

Earlier at HCL, I developed churn, engagement, and ticket-routing models using XGBoost, Random Forest, BERT, and TF-IDF, while optimizing SQL and PySpark pipelines supporting over 30,000 users. I focus on productionizing AI agents with observability, human-in-the-loop control, and continuous improvement.

Experience

Work history, roles, and key accomplishments

Current

AI Systems Engineer

Current

Quanta Manufacturing

Feb 2026 - Present (6 months)

Designed a multi-agent diagnostic system using Python and LangGraph that processed over 10,000 server logs to automate hardware validation with improved fault detection accuracy. Implemented scalable human-in-the-loop AI automation workflows using LangGraph and integrated tools like Cursor, Replit, and Claude to manage complex fault cases across 30+ concurrent systems.

Python LangGraph AI Automation Human In The Loop Cursor Replit Claude