The Data Architect Behind AI Prepared
Building the bridge between messy data and intelligent systems.
Most AI projects fail before a single model is trained. The data going in is broken, and nobody catches it until thousands of dollars are wasted. I built AI Prepared to fix that.
For the past 7 years, I've been deep in the data trenches: designing data structures for C-suite decision-making, building reporting systems from scratch, and leading teams through hands-on implementation. Teaching both front-end users and back-end developers how data flows is what separates an automated, accurate data architecture from one that just kind of works.
I watched companies pour millions into "AI Strategy" while ignoring the fundamental hygiene of their data. The pattern was always the same: great model, terrible inputs, confusing outputs.
I built this tool to solve the "Garbage In, Hallucination Out" problem. My goal is to give every engineer, product manager, and founder a fast way to audit their data before they waste tokens on training. Or worse, make decisions based on hallucinated outputs.
When I'm not coding, I'm writing deep-dives on LLMs, RAG architectures, and data topics. Usually through some nerdy analogy.
I'm always happy to jump in and help teams and organizations solve their data challenges. Reach out below if you have questions, need guidance, or just want to talk shop.
What I'm Building Now
My work on data architecture for AI led directly to JESTR — a tool that lets your team query your data warehouse in plain English, no SQL required. JESTR sits on top of your existing warehouse (Snowflake, BigQuery, Redshift) and routes questions through the right retrieval layer automatically.
If you're solving the “everyone has questions but only two people know SQL” problem, that's exactly what JESTR is for.