About
I build AI and data systems for businesses, and I test them well enough to say whether they are ready for real use.
I'm Kevin Street. I build AI and data systems for businesses, and I document how each one was tested, what it got right, what it got wrong, and what changed between versions. This site is my portfolio. The code and technical documentation for every project are on GitHub.
How I work
- Start with the outcome. Before any code is written, we agree on what the system has to do for your business and how we will know it is working.
- Prove it on your own examples. Early versions run on sample data that looks like your real work, so the results mean something to you.
- Keep people in charge. Any step with real consequences, such as sending, approving or changing a record, waits for a person to say yes, and every action is logged.
- Measure before you trust. Every system ships with a test set and repeatable checks for accuracy, made-up answers, failure handling, speed and cost, so you decide with numbers in hand.
- Leave it in good hands. The documentation covers setup, dependencies, data flows and test results, plus what is left before your own team can support the system without me.
Why testing comes first
A system that is right most of the time is only useful if you know which times. Before anything goes live, you get a plain table showing where it works, where it fails and what it costs to run. That is what makes it safe to put in front of your staff and your customers.
Let's Learn Out Loud
Separately, I teach what I build. Let's Learn Out Loud explains AI and data systems through real, open-source workflows: the chunking, embeddings, vector stores, model APIs and pipelines behind systems like the ones here, worked through step by step for people building their own.
Stack
Python · SQL · FastAPI · TypeScript · scikit-learn · Snowflake · AWS · Azure · Docker · Airflow · Git · Tableau · retrieval and vector databases · model APIs (Claude, GPT, Gemini) · model validation