I built and shipped a publicly deployed tender document extraction pipeline using the OpenAI Responses API, Structured Outputs, and Pydantic validation. My Tender Review Agent uses citation-grounded extraction and deterministic evaluation, achieving F1 0.889.
I also build RAG pipelines, local LLM inference workflows, and agentic automation with structured outputs and deterministic evaluation. My engineering experience includes construction estimation and BOQ workflows, where I’ve worked with tender documents, quantity take-offs, and project data.

