Founder Dossier
About Me
The builder's perspective behind AiBS and the standards guiding the product, data, and scope decisions.
ByColby Reichenbach
Overview
AiBS is built by one person working across product, engineering, and analysis.
AiBS is an independent product built by Colby Reichenbach. The work spans route design, frontend systems, data ingestion, database views, model audits, AI tooling, and the writing layer around the product.
That stack matters because the product is opinionated end to end. The interface, the models, the editorial framing, and the operational workflow all need to agree with each other. AiBS is not assembled as separate disconnected tracks; it is built as one system.
The standard behind the work is simple: claims should be inspectable, product decisions should be intentional, and technical sophistication only matters if it helps a user understand baseball more clearly.
Approach
The product and the codebase are meant to agree.
A recurring design rule in AiBS is that the public explanation should match the actual implementation. If the site says a model is audited, there should be scripts and artifacts behind that claim. If a page says a chart is AI-explained, the AI layer should be scoped and observable instead of hand-wavy.
That is why the product includes an unusually explicit About desk. It is not there as filler. It is there because part of the value of AiBS is showing how the product was built and what standards it is willing to defend.
Priority
Trust beats flash.
The same bias shapes product scope. Public AI is kept bounded to surfaces where the response can stay tied to a specific chart, game state, or analytic context. Query Lab and Copilot remain gated until they can meet a higher support and product bar.
That is the kind of bar I want the project to maintain. Better to ship a smaller product with coherent behavior than a broader product with blurry edges.