Hey HN — Jason here (Prepare 4 VC). We built Compass to help early-stage founders answer:
1) Where am I actually right now?
2) What should I do next week, specifically?
3) How do I keep that momentum every week?
How it works:
• 10–12 min diagnostic (“Chart My Course”) → five scores (Progress, Signal, Self-Efficacy, Execution, Market Awareness) + a momentum state
• Weekly reflections with an AI summary that turns notes into next steps + resources
• Optional layers: Navigator (monthly review), advisory (“Co-Pilot”), and fixed-scope execution (“Mission Control”)
Why we built it:
We run founder programs and advisory—too many tools, not enough operating rhythm. This is our attempt to make “assess → act → reflect” dead simple.
Stack/tech:
• Frontend: React (Lovable)
• Backend: Supabase (Postgres, Auth, Storage)
• Email: Resend
• AI: OpenAI via Edge Functions; hybrid scoring (deterministic rubric with small ±5 model adjustments)
• Auth: magic link with RLS; public resource storage via signed URLs
Privacy & data:
• We don’t sell data or train models on founder inputs.
• You can delete your account and data anytime from the dashboard.
What I’d love feedback on:
• Do the five scores + momentum state actually feel useful?
• Where would you want more transparency in the scoring?
• Would you prefer all-deterministic scoring, or is the hybrid approach acceptable?
How it works: • 10–12 min diagnostic (“Chart My Course”) → five scores (Progress, Signal, Self-Efficacy, Execution, Market Awareness) + a momentum state • Weekly reflections with an AI summary that turns notes into next steps + resources • Optional layers: Navigator (monthly review), advisory (“Co-Pilot”), and fixed-scope execution (“Mission Control”)
Why we built it: We run founder programs and advisory—too many tools, not enough operating rhythm. This is our attempt to make “assess → act → reflect” dead simple.
Stack/tech: • Frontend: React (Lovable) • Backend: Supabase (Postgres, Auth, Storage) • Email: Resend • AI: OpenAI via Edge Functions; hybrid scoring (deterministic rubric with small ±5 model adjustments) • Auth: magic link with RLS; public resource storage via signed URLs
Privacy & data: • We don’t sell data or train models on founder inputs. • You can delete your account and data anytime from the dashboard.
What I’d love feedback on: • Do the five scores + momentum state actually feel useful? • Where would you want more transparency in the scoring? • Would you prefer all-deterministic scoring, or is the hybrid approach acceptable?
Live demo: you can start here (free) → https://compass.prepare4vc.com If you’d rather read than click, here’s a 90s demo video: https://youtu.be/9gdqBh-1BWg
Happy to answer anything (tech, product choices, pricing, what’s broken). Thanks!