Ansh Upadhyay
@anshupadhyay
I build production LLM systems with RAG, evaluation gates, and hard cost controls.
IndiaWhat I'm looking for
I design and operate production LLM systems end to end, currently delivering around 48 exam-grade assessment sets each month to a paying CAT test-prep client.
My six-stage pipeline orchestrates Anthropic Claude, OpenAI, and Google Gemini, with provider switching, structured evaluation, compliance gates, human review, and per-request cost caps. I built a pre-call guard that prevents calls exceeding tier-based spend limits and rejects 26% of candidates before API cost.
I built RAG retrieval with LangChain, ChromaDB, and sentence-transformer embeddings, plus a nine-channel novelty auditor that eliminated near-duplicate delivery across 233 logged audits. I also shipped a FastAPI operations console for live jobs, queues, and corpus health while preserving read-only database access.
Previously at EY India, I built Airflow and BigQuery pipelines for audit-grade banking and insurance reporting for HSBC India and Aditya Birla Sun Life. I improved insurance reporting efficiency by about 80% and reduced manual data-cleaning effort by about 40% through Python automation.
Experience
Work history, roles, and key accomplishments
AI Engineer & Founder
Freelance
Jul 2026 - Present (2 months)
Design and operate a six-stage production LLM pipeline generating exam-grade reading-comprehension assessment content, shipping ~48 approved sets per month to a paying institutional client. Engineered cost guardrails and provider-agnostic orchestration across Anthropic, OpenAI, and Google Gemini.
Built and operated Apache Airflow ETL pipelines for regulatory reporting at HSBC India, ensuring audit-grade output under external review. Designed data-quality gates, tuned BigQuery SQL, and delivered insurance policy reporting for Aditya Birla Sun Life, improving efficiency by ~80%.
Education
Degrees, certifications, and relevant coursework
Lakshmi Narain College of Technology
Bachelor of Technology, Computer Science
Grade: 7.8/10
Completed a Bachelor of Technology in Computer Science in 2023 with a grade of 7.8/10.
Availability
Location
IndiaAuthorized to work in
Website
github.com/Crescent-1Salary expectations
Social media
Skills
Interested in hiring Ansh?
You can contact Ansh and 90k+ other talented remote workers on Himalayas.
Message AnshGet matched with your dream remote job
Sign up now and join over 250,000+ remote workers who receive personalized job alerts, curated job matches, and more for free!
