AI/SaaS MVP Development with CRM

need help building MVP for AI/SaaS app.

Core MVP Features
1. Call Recording Ingestion
- Integrate Zoom, Google Meet, and Microsoft Teams APIs
- Capture call audio and metadata
2. Transcription Layer
- Use AssemblyAI or OpenAI Whisper to transcribe calls
- Speaker separation if possible
ClosePilot.AI - MVP Job Post (MVP GOOD)
3. AI Coaching Layer (GPT-4o)
- Analyze discovery strength, objection handling, talk vs listen
- Output coaching notes and follow-up messages
- Sentiment and buyer intent analysis
- Coaching aligned to MEDDIC, Challenger, SPICED, and consultative best practices
4. User Dashboards
- Rep dashboard: call history, performance scorecards, generated follow-ups
- VP dashboard: team trends, flagged issues, buyer interest signals
5. Light CRM Integration
- HubSpot or Salesforce sync (log summaries, attach insights to deals)
6. Auth + User Roles
- AE, VP/Sales Manager access
- Firebase or Supabase for backend/auth
Tech Stack (Preferred)
- Frontend: React (or Bubble for low-code option)
- Backend: Node.js or Python (FastAPI)
- Transcription: AssemblyAI or Whisper
- AI Layer: OpenAI GPT-4o
- Storage: Supabase, Firebase, or PostgreSQL
- Optional: Zapier or Make for workflow automation
Deliverables
- Fully functional MVP hosted on Vercel, Render, or similar
- Clean GitHub repo with documented code
- Live user interface for basic demo (AE and VP roles