AI engineer. RAG systems, fine-tuning small open models, and the evals that prove either one worked. First-author paper under review at ICLR 2027 on whether LLMs hold instructions under pressure. Shipped a speech-to-speech assistant live to 70+ training centres.
https://freelance-gamma-bice.vercel.app
eshan.worke@gmail.comAvailable for freelance, fixed price:
Cited Q&A over your docs - $900 / 10 days
Frontier API call -> fine-tuned 1-8B model - $1,200 / 2 weeks
Eval suite for an LLM feature you shipped - $700 / 1 week