The product
Four surfaces. One substrate.
The four items shipping in V1.
01 / interview brief generator
Upload. 90 seconds. Ship a roadmap on it.
- Accepts Zoom, Meet, Loom, Granola, Otter, Fireflies. Or raw audio. Or manual entry.
- Four-dimensional tagging pass. Says diarised. Thinks extracted. Does observed. Feels flagged.
- Value-risk synthesis as three evidence-linked bullets. That is the headline output.
Guarantee: under 90 seconds render, p95. On Lagos 10Mbps upload.
02 / shareable brief URL
Every brief is a link. Every link travels.
- Public by default. Auth-optional viewing. Workspace/team/private toggles.
- One-click Slack, WhatsApp, LinkedIn, X, email share.
- Open-Graph and Twitter-Card unfurls with the Value-risk synthesis as preview.
Guarantee: 5 minutes from signup to a brief shared in Slack, p95.
Bulk payments discovery · 38 min
She will choose this if bulk payments land under a 1.5% effective fee.
Slack unfurl preview
RICS Brief · Bulk payments discovery
Value-risk: chooses under 1.5% effective fee. 3 evidence-linked bullets.
03 / brief corpus query
Every insight you ever captured. One query away. Forever.
Cross-brief synthesis · 3 briefs matched
Segment leads converge on a fee ceiling between 1.2% and 1.5% effective. Two of three flagged treasury approval as the blocking step above ₦5M. Willingness-to-pay sits in Thinks, not Says, in all three briefs.
brief #7 · Segment Lead A · Mar 12
Says[19:47] "anything above one and a half points is a hard no" · Thinks[19:52] threshold is effective rate, not sticker
brief #4 · Segment Lead B · Feb 28
Says[22:10] "we benchmark against our current 1.2%" · Thinks[22:31] anchor is the incumbent, not value
brief #2 · Segment Lead C · Feb 14
Says[33:18] "legal reviews anything over $15 a seat" · Thinks[33:29] procurement threshold is the real ceiling
- Natural-language queries against the entire brief corpus.
- Cross-brief synthesis when a query matches more than one brief.
- Faceted filters: date range, participant, segment, risk tag, empathy dimension.
Guarantee: under 3 seconds to first result, p95.
04 / MCP server
Every AI agent in your team's stack, reading the same substrate.
- Read tools day one: list_briefs, get_brief, query_briefs, get_empathy_field, get_risk_tags.
- Write tools V1.5: emit_brief, update_empathy_field, set_risk_tag.
- Workspace-scoped API keys. Rotation and revocation in-console.
Guarantee: under 5 seconds MCP tool-call latency, p95.
bonus / free surface
Not shipping yet? Take the free audit.
Ten questions. Two minutes. No signup. A scored result across the 4x4.
Take the audit →