Looking for a job

Devesh User

@deveshuser

Message

I build reliable LLM systems, RAG pipelines, and AI backend services.

India

Message

What I'm looking for

I'm looking for an AI engineering role where I can build reliable LLM applications, RAG systems, and backend services while growing through hands-on projects, mentorship, and collaborative technical review.

I'm building LLM-powered backend systems and AI agents that turn complex inputs into reliable, useful outputs.

At ATGWorld Networks, I built the FastAPI-based Interview Evaluator for Vidya AI, using Google Gemini to convert interview transcripts into readiness scores and skill breakdowns. I also integrated Pinecone for transcript retrieval and persistent evaluation storage.

I traced and fixed a critical data-loss issue across model output, schemas, and backend payloads, restructuring validation for reliable data flow. I also designed a tiered backend engineering roadmap built around real projects, AI pair programming, and human-in-the-loop review.

My projects include Cascade AI, a multi-agent orchestration system with persistent Firestore memory and self-healing fallbacks; a GAIA benchmark agent that achieved 60% accuracy; and a research-paper RAG chatbot grounded in Qdrant-stored semantic embeddings.

Experience

Work history, roles, and key accomplishments

Current

A.I Developer Intern

Current

ATGWorld Networks

May 2026 - Present (4 months)

Built and integrated a FastAPI backend service for Vidya AI's interview platform using Google's Gemini AI model to convert interview transcripts into structured evaluation data. Diagnosed and resolved a critical data-loss bug and designed a tiered backend engineering training roadmap.

fastAPI Pinecone Python Backend Development Data Validation