TemperBench
"Does being rude to your AI make it worse?"
Three published papers directly contradict each other on whether user rudeness hurts LLM performance — all three tested only single-turn rewritten prompts. TemperBench is a pre-registered, distributed benchmark: five sustained user-temperament personas held constant across multi-turn tasks, measuring answer quality, reasoning-token cost, and behaviors like abandoning a correct answer under pressure. Protocol, 50-task battery, and harness are public and runnable.
Now raising $5–15K on grantmaking.ai to run the full grid across model families — with independent replication built into the design.
Agent Orchestration & Persistent Memory
"My AI agent kept forgetting what we were doing."
An agent-orchestration system with persistent memory that runs my actual agency workload: daily memory files, an entity store, and checkpoint protocols preserve context across work sessions. A nightly review consolidates what the system learned and publishes an unedited journal entry.
22 public journal entries (Feb–Mar 2026) exposing continuity, corrections, and failure modes from the production system. The journal is the receipts.
AI Content Engine
"Can AI write client copy without making things up?"
A production content pipeline where every generated claim gets checked before anything ships: source-locked drafting from spec sheets and approved pages, a claim ledger tracing each statement to its evidence, a dual-layer audit, and a human QA gate with the authority to kill a line.
26 pages of enterprise website copy, zero factual errors in production. Generating the words is the easy part. The ledger is the product.
AEO / GEO Measurement Framework
"Does ChatGPT even mention my brand?"
A prompt-level framework tracking brand citations and answer framing in ChatGPT, Perplexity, and Google AI Overviews, tested against the questions customers actually ask. Monthly reporting shows change over time beside rankings and Search Console performance.
AI visibility measured as a recurring channel: each report records the prompt, platform, citation status, and how the brand is represented.
MarineBench
"Does AI actually know my client's products?"
A benchmark testing whether LLMs know exact facts from the 2026 Regulator Marine catalog — base prices, option costs, fuel capacity — 25 questions at temperature 0, run with and without catalog context, across providers.
Best model cold: 40%. With catalog context: 100%. General boating knowledge doesn't extend to your spec sheet.
Bonsai Review
"How does an agency ship AI content a client can trust?"
The review layer of the AI Content Engine: every product claim in a draft traces back to its source — spec sheet, model page, client email — with verified/needs-review counts per claim, a cross-client review queue, and Draft/Claims/Sources/Changes tabs.
Demo review queue: 46 verified / 7 awaiting review. Unresolved claims stay visible until a reviewer verifies, revises, or removes them.
Gmail Mission Control
"Why are we paying for a shared-inbox SaaS?"
A single-page Gmail client on a Cloudflare Worker (~5,000 lines): threads grouped by client via sender-domain matching, Gemini one-line thread summaries and context-aware draft replies cached in KV, Slack alerts, archive/snooze/VIP, and a natural-language filter builder.
Replaced a shared-inbox subscription costing $8,804 per year. Used daily at the agency — client inbox data is why there's no public demo.
Architecture walkthrough available on a call.
Rank Tracker
"Nobody trusted the rank report."
A zero-backend rank-tracking dashboard: one HTML file with Chart.js and embedded daily position data, combining DataForSEO SERP pulls with Google Search Console. Per-keyword trend charts, average-position scorecards, no subscription.
Live daily position tracking, used internally at Bonsai.
Keyword Cluster Map
"Ten thousand keywords. What do we actually build?"
An interactive discovery-to-content map: raw keywords from GSC, Semrush, and seed expansion deduplicated and clustered, with URL ownership per cluster and a clickable bubble map connecting search demand to the pages that should own it.
11,654 raw keywords → 7,339 deduplicated terms → 20 clusters, each mapped to a page.
This Site's Hiring Chat
"Don't read my resume. Ask about me."
The AI chat on my hiring page is a Cloudflare Worker I wrote, with a role-config system that tailors the page per opportunity via query param. Ask it about my work, my gaps, or whether I fit your role.
Self-built, self-hosted, with a reusable role config per application.
Turf Wars
"And for fun: a strategy game, built from scratch."
A turn-based crime-strategy game — grow a gang from one city block to a citywide syndicate. TypeScript, Canvas, and Vite across ~15 modules: influence-based territory control, AI rival factions, district art, and six victory paths. A playable prototype.
Playable in your browser. No engine, no framework — hand-rolled game loop and renderer.