Turn every demo call into structured data three teams can use

Retrieve the Demo meeting type from the meeting taxonomy. Confirm it exists with notetaker autojoin enabled. If absent, halt and direct the user to /meeting-types-taxonomy.

Configure the AI summary recipe for the Demo meeting type. Extract structured fields: (1) Features Demoed — list of product features shown, with engagement signal per feature (attendee asked questions / silent / pushed back); (2) Questions Asked — list of attendee questions with category (capability, integration, pricing, security, onboarding, other); (3) Objections — list of objections raised verbatim, with category (price/timing/competition/feature-gap/authority/trust) and resolution status (handled/parked/unresolved); (4) Technical Concerns — specific technical questions or blockers (integrations needed, data residency, SSO, compliance); (5) Decision-Maker Signal — was a decision-maker on the call, were they engaged; (6) Follow-up Requested — what attendee asked for (case study, trial, technical demo, custom proposal, references); (7) Next Step — what was agreed. Use 'not_discussed' for missing fields rather than guessing.

Compose a demo insights dashboard reading from structured summaries: top 10 features by demo frequency (which features sell themselves vs. need more pitch); top 10 objections by frequency (objection-handling content priorities); decision-maker attendance rate (low % = multi-threading problem); follow-up-requested distribution (what does the market actually want); and most-asked technical concern categories (product / engineering input). Group by rep and time period.