AI / POC Lab
Not Just Aware of AI. Using It.
This section documents what I've actually built, working systems applied to real problems. Not demos for the sake of demos. Each project started with a genuine need, got built across multiple sessions, and is either in active use or in progress. The stack varies. The approach is the same: figure out what AI can actually do by building something that has to work.
Anthropic ships constantly: Claude Code releases, SDK updates, API changes, new models, model retirements. Most of it doesn't matter to any given project, but some of it breaks things. I wanted to know which was which without reading every changelog.
- 1FetchNew items from eight Anthropic sources, on a schedule you choose or on demand.
- 2ReviewThree tool-less Claude calls judge impact, kind of change and risk; code merges them into an urgency level.
- 3Check your filesThe names in each item are looked up locally in your project folders and Claude config.
- 4RateCode bounds the 0 to 10 range per project; a final no-tools call places the item inside it.
- 5BriefTop picks become one-step briefs in each project, with findings asked for before any change.
Most families don't know how to build a balanced college list. They navigate with spreadsheets, gut feel, and conflicting advice from counselors. AdmitKit asks the right questions, calculates personalized admission probabilities, surfaces student perspectives from YouTube, and recommends a list balanced across Safety, Target, and Reach schools. The tool that started as a custom build for one family is now a public beta available to every family.

The engine is built and heavily proven: eight of eleven build blocks complete, all 31 proofs green, re-run twice to show repeatability. Discovery, Strategy and the Go/No-Go pair are producing. The next build is the gate screen where a person makes the call.
- 1DiscoveryA brief goes in; an agent produces a structured brief plus a set of facts, each one declaring whether it is sourced or unverified. Cost is accounted at four rates from the first call.
- 2Review and dissentA reviewer from a different model family iterates the output to a quality bar, with a premise-gated stop that fails closed rather than pushing a weak premise forward.
- 3Opportunity screenThe idea is screened against a per-product cost baseline, so the spend floor for pursuing it is known before anything expensive runs.
- 4Bull and bearTwo agents argue the Go/No-Go, one for and one against, with the database enforcing that they come from different model families.
- 5Sealed packetThe arguments and their evidence are frozen into a tamper-evident packet, so the decision can later be audited against exactly what the human saw.
- 6The human gateA person reads the packet and decides. Nothing irreversible happens without that signature, which is why the gate screen is the next thing being built.
Phase I complete. Phase II active. Stakeholder intelligence shapes agent decisions and tailors BRD/PRD exports to individual audiences. Seven documented runs across ERP forecasting, e-commerce, athletic training, furniture making, and professional networking, with 100% schema compliance since Round 2. The pipeline adapts naturally to non-software domains without special handling.
Most senior executives either pay an agency and wait weeks for a professional site, or settle for a generic template that undersells them. The goal was a founder-forward dual-audience site with real content and a live AI/POC Lab - without the agency timeline or cost.

Just for fun2 satirical baseball sites
Built with the same care as the serious work: live data, and formulas you can check.
Traditional baseball projections like PECOTA are statistically rigorous and emotionally empty. Fans know a season is not just wins above replacement. It is tabloid drama, mascot charisma, and whether the bullpen can find the strike zone in August. This project puts those variables in the equation.

Hot takes on a big contract never have to be right. This site commits to its rule first, publishes the thresholds, changes the verdict only when the evidence holds, and keeps every past verdict on the page with its date.

