Kendal · CoWork · 2025 to now
HA note I wrote myself, the night it started
CoWork is the AI assistant I designed for individual real estate brokers. It started as a 24-hour demo and became the platform 200+ brokers pay for.
01 · The problem
In every Kendal CRM demo, brokers nodded at the dashboards, then scrolled WhatsApp to find yesterday's lead. Their day lives in chats, calls and voice notes. The CRM is the one place it doesn't.
| Contact | Stage | Last update |
|---|---|---|
| Omar Haddad | — | 34 days ago |
| David Chen | — | 21 days ago |
| Sarah Chen | — | never |
Built for the agency's reporting. Fed by nobody.
Where the deals actually happen.
Agencies loved Kendal. Brokers kept asking, "what about me?"
Kendal's AI already qualified leads around the clock for agencies on WhatsApp, more than a million of them. But the individual broker, the person actually closing, still started every morning from zero.
02 · The insight
Before a broker can sell anything, she answers the same three questions every morning. If the product answered them when she woke up, she'd never need to open a CRM.
The viewing in 58 minutes, the offer that expires at noon, the prep she forgot.
Who called three times, opened the brochure, or went quiet after a viewing.
The floor plan she promised, the reply she meant to send, the comps for a seller.
A CRM asks brokers to feed it. CoWork reads her inbox, WhatsApp and calendar, and feeds her.
Sat 03:04A text from Joel.Mon 10:00The pitch.
03 · Her day, built
This is the real CoWork, recorded at desktop size on demo data. Nothing here is a mockup. Each moment plays where she'd be at that hour.
The daily cover greets her by name and counts what's ahead: ten events, nine AI to-dos. One quick action drafts follow-ups for everyone who needs one. Scroll down and her hot leads and a timed agenda follow, viewings marked in amber.
Playing: the morning dashboard, top to bottom.Every lead carries its reason. Harman Bassi called three times in two days, completed two viewings and has the brochure, and each signal names its source: call log, calendar, email. The next move is right there: call today to confirm pricing and close.
Playing: opening Harman's lead and its signals.Adding a contact starts the mic, so she just talks. "So Elias called back, he's now looking at AED 9.5 million, wants to close within 2 months" turns into a budget and a timeline, each waiting for her tick. Nothing is saved until she approves it.
Playing: a new contact, mic already listening.She walks the property once. CoWork pulls a sharp, straightened still for every room, applies her enhancements and writes the title, description and details from her voiceover. Nothing is published until she has reviewed every photo and field.
Playing: walkthrough to finished listing in 40 seconds.WhatsApp, inside CoWork. The AI brief already knows David is negotiating the Palm Jumeirah villa, AED 16.5M against an 18M ask, and wants a second walkthrough with his wife. She can speak a reply, let AI draft one, or rewrite her own, then send it herself.
Playing: AI drafts David's reply.04 · The outcome
The agency saw a broker's morning sorted before she'd opened WhatsApp. So the demo stopped being a one-off and became how Kendal sold.
05 · What changed
The demo waited for a prompt. A broker at 8am doesn't know what to ask, so the platform answers first and keeps chat for follow-up questions.
The demo ran on fixtures. The product reads her actual inbox, WhatsApp and calendar, plus new leads from the portals, so the brief is about her morning, not a sample one.
Same feature, local language. A CMA in Dubai is a Valuation Report in London and a Land & Depreciation valuation in Tokyo, with each market's own portals and currency.
What I'd do differently
No design system was right for 24 hours. It stopped being right the day "CMA" had to become "Valuation Report" everywhere at once. Next time, I'd name the tokens before the second market, not after the third.
Next case study
Toddle Insights →Product screens are rebuilt from CoWork's code with demo data. People, listings and prices inside them are fictional, and Aisha is a composite of the brokers we met.