A working prototype

For Jacinta.

June 2026 · built from your LEAP session

isonew · LEAP follow-up · ReCity Durham cohort

WholeView
Intake Engine v0

At LEAP you walked me through your week, step by step: book the call, log the contact in a sheet, run the conversation, drop the transcript in Claude, hand-build a report, send the email. ~10+ hours, every week. This is that pipeline — running.

Built by Ronan Pinho · for Jacinta Walker · Equal Pay Project / WholeView Report

What you told me, in your words

I built this off your answers — not a template.

"My weekly workflow: booking customer discovery sessions in my Google Calendar; then adding their name and contact info in a Google worksheet; then pulling data and questions to ask them; then having the conversation; putting the transcript in Claude; printing out a report — summary, next steps, research items, follow-up email, and action items; sending a Google email. This process takes about 10+ hours weekly." — Jacinta, LEAP question 4 · 2026-05-13
"I don't want to be the only person knowing what to do and how to do it. This needs to be a system." — Jacinta, LEAP question 10 · the 12-month picture
10+
hours/week on this pipeline — your number
$200
your hourly value (top bracket you picked)
~$104K
/year of your time, conservatively (10h × $200 × 52)

Run it · part 1 — the meeting-to-action pipeline

Paste a discovery transcript. Get your report.

This is the exact output you hand-build today: summary · next steps · research items · a ready-to-send follow-up email · action items. Paste a real Fathom / Read.ai / Claude transcript — or hit Load sample — and watch the hour collapse into a few seconds. Nothing leaves this page; it runs entirely in your browser.

Your report will appear here — summary, next steps, follow-up email, and the minutes + dollars you just got back.

Run it · part 2 — the intake you chose in the room

Four scattered tabs → one canonical intake.

In the session you picked Intake Form Automation over the email router — because the real mess is the data, not the inbox. Today your records live in separate Google Sheets tabs by stakeholder type. The engine reads all of them and lands every record in one schema, tagged by who they are, so you (and your incoming intern) work from a single source.

Educational institutions→ Employers→ Workforce dev→ Advisors ⟶one canonical intake
// sheet tab: educational_institutions
name · contact · school · notes…
// sheet tab: employers
company · contact · role · notes…
// sheet tab: workforce_dev
org · contact · program · notes…
// sheet tab: advisors
name · expertise · contact · notes…
canonical_intake
id · uuid
full_name
stakeholder_type · [institution · employer · workforce · advisor · beta]
contact · email / phone
source · discovery / referral / pilot
stage · prospect → discovery → follow-up → beta
last_touch · auto from calendar + email
next_action · auto from report
context_pack · the warm handoff for whoever picks it up

This is the CRM you started in Airtable and stopped — minus the learning curve. It captures discovery, follow-ups and beta testers, exactly what you said you needed, and runs in the stack you already pay for.

From prototype to your stack

This page is the demo. Here's the real thing.

You're already in Claude Code + Lovable + Supabase — so you know a prototype isn't a system. The version that reclaims hundreds of hours runs inside your Google Workspace + Airtable, in short tested chunks, the way you asked at LEAP.

Wire the pipeline to your transcripts

Fathom/Read.ai transcript in → report out → draft email in your Gmail → record updated. The part you do by hand, automated, in week one.

Stand up the canonical intake

The four tabs collapse into one schema in Airtable (or Sheets, your call), with stakeholder tags and the context packs your intern can run on.

Hand it off as a system

Documented, with check-ins — so it's not living in your head. That was your 12-month goal; we just start now.