2.7 KiB
2.7 KiB
PM Brief — AI Journaling → LinkedIn (1‑page)
Slide purpose
- Quick, shareable one‑page brief for the PM to review before the meeting. Includes the demo script, current status, key decisions, risks, and next steps.
Project one-liner
- Mobile-first journaling app that converts daily text/voice entries into LinkedIn-ready posts using Whisper (transcription) and OpenAI (generation). Local-first by default; direct LinkedIn publish supported.
Current status (what we have)
- Persona & mind map captured: Senior PM at mid-size SaaS.
- MVP scoped (text + voice capture, transcription, OpenAI-based conversion, draft editor, LinkedIn publish + fallback).
- Sprint tickets (P0–P2) and serverless API spec prepared.
- Sample post conversions (3 entries × 2 variants) ready for demo.
Demo script (10–15 minutes)
- 30s — One‑line context and persona.
- 2 min — Capture: Quick text entry + optional voice record (20s). Show local-first saved entry.
- 1 min — Transcribe: upload → Whisper → show editable transcript + anonymize toggle.
- 2–3 min — Convert: call OpenAI → show 2 variants, hashtags, CTA. Edit variant in-line.
- 2 min — Publish: Direct LinkedIn publish (if connected) OR fallback: Copy & Open LinkedIn (native share).
- 1 min — Analytics & logs: show counters and processing consent logs.
- 30s — Decision checklist and next steps.
Key decisions requested from PM
- Retention default: 30 / 90 / 365 / indefinite?
- LinkedIn publish default: opt-in (recommended) or auto‑publish?
- Monetization: free capped usage vs paid tier at launch?
- Consent for optional telemetry (to monitor API costs)? Yes / No
Budget & timeline (reminder)
- Budget: $5,000 for API, hosting, storage.
- Timeline: 3 months to MVP (single operator). Prioritize P0 tickets in sprint checklist.
Top risks & mitigations
- API cost overrun — enforce quotas, use cheaper models, limit variants.
- LinkedIn API delays — use fallback share flow; use test account for demo.
- PII leakage — anonymize toggle, redact UI, local-first default, ephemeral audio deletion.
Immediate asks for PM meeting
- Approve retention default and LinkedIn publish policy.
- Approve budget allocation and monitoring thresholds.
- Nominate owner for LinkedIn app registration (or approve me to register test app).
- Confirm preferred demo time and attendees.
Contact & attachments
- Owner: Babz (operator)
- Attachments:
SPRINT-TICKETS.md,SERVERLESS-API-SPEC.md,PROMPT-TEMPLATES.md,DEMO-INSTRUCTIONS.md(inbmad/bmm/agents/hand-off/)
Next step
- Run the live demo with PM and convert decisions into assigned GitHub issues from
SPRINT-TICKETS.md.
(Prepared for PM review — adjust owner names and meeting time before sending.)