Core thesis: The Algo Academy isn’t an education platform that uses AI — it’s an AI platform that educates. Every surface is powered by conversational agents, not bolted-on chatbots.

Buy vs. Build strategy: Build the agentic layer (that’s the moat). Buy the commodity infrastructure — Clerk for auth, Stripe for payments, Mux/Bunny for video, Stream for chat, Vercel Postgres for data, PostHog for analytics.

7 Platform Agents:

  1. Onboarding Agent — walks new students through setup and CLAUDE.md creation
  2. Curriculum Agent — adaptive, progress-aware guide (not static modules)
  3. Implementation Agent — pair programmer for agent deployment
  4. Accountability Agent — proactive nudges and milestone tracking
  5. Community Agent — thread summaries, student matching, Q&A
  6. Assessment Agent — readiness scoring at Day 30/60/90
  7. Strategy Agent — Operator tier exclusive ($497/mo), personalized business advisor Monetization layers: Subscriptions ($47-$497/mo), usage-based API overages, an agent template marketplace (70/30 split), and future enterprise/white-label.

5-phase build sequence: ∙ Phase 1 (Weeks 1-6): MVP with auth, dashboard, onboarding + curriculum agents ∙ Phase 2 (Weeks 7-12): Full agentic layer with all 7 agents ∙ Phase 3 (Weeks 13-18): Community platform ∙ Phase 4 (Weeks 19-26): Marketplace ∙ Phase 5 (Month 7+): Enterprise and white-label

The retention moat: Not the curriculum (that can be copied) — it’s the student’s accumulated agent context. Their CLAUDE.md, workflow commands, performance data, and agent configs all live in-platform. Leaving means starting over.

381 lines, well-structured. This plus the 90-day curriculum are a strong foundation.