AI Engineer and Technical Product Manager. I ship production GenAI — agents, RAG, NL2SQL, evaluation — and own the product decisions around it. Two years, at a startup small enough that both were the same job.
The products below are my employer's; the engineering and the decisions are mine to describe.
Designed a tool-calling layer end to end — how a task is decomposed, which tool runs, what happens when one fails, and where a human is put back in the loop.
Ingestion from documents, URLs and raw text; chunking strategy chosen against a benchmark rather than a default; embedding-model selection; retrieval tuned on real queries.
Schema-aware prompting, query generation and validation, read-only execution, streaming and structured output. Built the pipeline and owned the product through public launch.
Fixed benchmarks with recorded provenance per run — commit, prompt version, model, post-processing version — so a change can be attributed rather than guessed at.
Python services behind auth — streaming responses, webhooks, background workers, caching and usage limits. Containerised and deployed on AWS.
Supervised models taken the whole way — exploratory analysis, feature engineering, training, and evaluation against a held-out baseline. Regression on retail sales history, classification on product data, experiment runs tracked so a result can be reproduced.
Specs with screen-by-screen UI to API mapping, validated as clickable prototypes before build. Acceptance criteria on every task, scope and launch calls, and the decision about what not to build.
The two bars are the part most demos skip, and the part that decides whether a system survives contact with real users. I have built both.
The habit underneath all of it: I measure before I choose. A model swap proves itself on accuracy against a fixed benchmark before cost enters the conversation.
Open source, and mine. New repositories appear here on their own.
I trained a small AI on 5,851 correct database queries and measured it against an untrained copy of itself. It got better at the task and no steadier at it — and the training quietly turned crashes into wrong answers that raise no error at all.
Medical Q&A grounded in a PDF corpus. The hard part is not retrieval — it is getting it to say "my sources do not cover that" instead of improvising.
Ask an AI the same database question ten times: do you get the same right answer? 14,940 AI-written queries show a 13.7-point gap between what a model can do and what it does every time, and that most of the main model's mistakes raise no error at all.
Five parsers, one harness, the same document — PyPDF, PDFPlumber, PDFMiner, LlamaParse and AWS Textract, side by side. The comparison is the artifact.
A habit coach where the model never does arithmetic. Streaks and rates are pure Python with fifteen tests; the LLM only interprets the finished summary.
Semantic product search with price and rating filters, served over FastAPI.
Daily sales across 1,115 stores — feature engineering and regression model comparison.
Everything below is something I have actually built with. The terms teams ask for most lead each group.
Which model and vendor, RAG or an agent loop, what cost and latency the accuracy is worth, and how exposed a choice leaves us when a provider deprecates it.
Owned AI products end to end while building them — agent capability from prototype to production, NL2SQL, retrieval systems, and the evaluation and observability layers underneath.
Wrote the specs and acceptance criteria the team built against, made the model and architecture calls, and led a cross-functional team of six through to public launch.
Six-month full-time internship on the team's LLM work — retrieval pipelines, prompt and evaluation experiments, and the Python services around them. It led to the full-time offer.
Most people in AI are the engineer or the one deciding what to build. I ended up both, at a startup small enough that the title covered the job rather than describing it. I decide what to build and what not to, turn it into specs engineering can work from, and build the parts I specify — which is why my feasibility calls hold up in the room.
What I want next is more depth on production AI: evaluation, orchestration and reliability as the job rather than the afterthought.