A 6-week cohort that took participants from AI fundamentals to building real systems — with zero paid marketing and zero cancellations.
The AI Architect Program was a community-run workshop series (December 2025–February 2026) designed to bridge the gap between AI curiosity and AI competence. Seven sessions, 91 registrants, and a participant base that spanned Fortune 500 companies, biotech, manufacturing, academia, and independent founders — all organic, all word-of-mouth.
Sessions were opt-in and structured based on content participants surfaced during the Orientation Session "Session 0": LLM fundamentals → prompt engineering → AI workflows → vibe coding → AI agents → multimodal AI → a final Show & Tell showcase where participants demoed what they built.
| Metric | Result |
|---|---|
| Registrants | 91 (zero paid marketing) |
| Satisfaction rate | 96% |
| Avg helpfulness rating | 4.47 / 5.0 |
| Post-program overall rating | 4.8 / 5.0 |
| Season 2 continuation intent | 100% |
Each session generated three data streams:
- Attendance CSVs — unique attendees per session, tracked against registration list
- Post-session polls — helpfulness ratings and open-ended feedback after each session
- End-of-program survey — overall satisfaction, NPS-style continuation intent, expansion topic requests
All data was collected manually, anonymized, and stored as flat CSV files.
Raw CSVs → structured insight, fast.
Rather than manually aggregating and formatting metrics, I used Claude Code to:
- Analyze attendance patterns and calculate retention rates across 7 sessions
- Aggregate satisfaction scores and surface qualitative themes from open-ended responses
- Generate
generate_reports.py— a Python script that reads the raw data and renders the full HTML report with all charts and metrics
The result: a polished, stakeholder-ready data story produced from raw CSVs in a single session. The report is live at AI-Architect-Program-Report.html.
This repo is itself a demonstration of the program's core thesis: AI tools are most powerful when you understand how to direct them with intention.
-
1beebe/claude— A live artifact from Session 6 (Show & Tell). Demonstrates the LLM + tools + skills architecture taught throughout the program: the same model and tools produce measurably better outputs when guided by domain-specific knowledge. -
1beebe/ai-agent-portfolio— AI Agents & Enablement Portfolio. Production-ready agents and workflows in customer education, content automation, and applied AI for learning design.