anshmangukiya

Location: Surat, India (IST — overlaps US mornings and EU afternoons)

Remote: Yes

Willing to relocate: No

Technologies: Python, FastAPI, LangGraph, LangChain, vLLM, QLoRA/PEFT/TRL, Hugging Face, Qdrant, Pinecone, pgvector, PostgreSQL, Docker, multi-GPU Linux, .NET/C# integration

Résumé/CV: https://anshmangukiya.in

Email: [email protected]

SEEKING FREELANCE / CONTRACT WORK. AI engineer building production GenAI systems at a US software company. Freelance: assistants on your own documents with cited answers, self-hosted open-weight LLMs on your own GPUs, and QLoRA fine-tuning when retrieval isn't enough. Production numbers: 70% lower support load from one LangGraph agent, 80% less manual integration code, 150+ tokens/sec inference.

Try it: a RAG assistant on my own work, with sources — https://anshmangukiya.in

Write-up: 332 fine-tuning runs on consumer GPUs, every failure included — https://anshmangukiya.in/fine-tuning-field-notes

Fixed-price projects preferred; before we talk price I'll build a demo on your documents.