Proposal · for CTO (Matthew Chan) and CPO (Aaron Kurzak) review
One day in Hong Kong with 20 Crypto.com colleagues, building ai.com with us
Individual contributors, half new to ai.com and half from the alpha, show us where it works, where it doesn't, and what they'd build instead. The morning is testing. The afternoon is co-creation.
01Why
Most of ai.com's direction so far has come from inside the team. Before we commit to a roadmap, we want to watch real first use, hear what people's work actually needs, and let them shape the product rather than just pick between our options.
- ICs only. People speak more freely without managers in the room, and their needs differ. Managers and execs get a separate session later.
- No pitch. We don't present a vision. Several possible directions, ours included, go in front of people unlabelled, and the evidence decides.
- Real data. Some participants connect their work tools the week before, so they test ai.com on their own work, not a demo.
- Quiet voices count. Every idea round starts with silent writing, then pairs, then small groups.
02What we'll come away knowing
| Q1 · Needs | What do Crypto.com ICs struggle with day to day, and how do they use AI today? |
| Q2 · Futures | Which of six possible futures would they want tomorrow, and why? |
| Q3 · Shell | What to keep, cut and steal from Jeet's and Aaron's concepts, plus a "Concept 3" from participants. |
| Q4 · First run | What frustrates and delights someone using ai.com for the first time? |
| Q5 · New ideas | What haven't we thought of? |
| Q6 · Priorities | What matters most when their ideas compete with ours, and what would make them switch tools? |
03Who's in the room
20 ICs in two breakout groups of 10. Each has 5 Alpha Users and 5 who have never used ai.com (Red: 2 Fresh, 3 Early-setup. Blue: 3 Fresh, 2 Early-setup). At least half come from non-engineering functions, and nobody has direct reports.
| Group | Who | Before the day | On the day |
|---|---|---|---|
| A · Fresh eyes · 5 | Never used ai.com | A survey only. No access. | First use, live and observed |
| B · Set up early · 5 | Never used ai.com | Sign up and connect their work tools this week, by Wed 30 Sep | Ask it about their own work |
| C · Alpha users · 10 | Used it in the alpha (from the alpha user list), including some who stopped | A survey | What works day to day, and what they gave up on |
Our side (5): Jeet facilitates. Sunny and one engineer moderate the two breakouts from scripts, and two engineers take notes, with Scribe as a cross-check. Jeet doesn't moderate either concept. Aaron isn't attending but reviews the materials beforehand.
04The day
Cyan = testing, green = creating. A 15-minute kickoff opens the day, a survey closes it, and the team debriefs until 16:30. Full timings: Run of show.
05What comes out
| Day +2 | Evidence readout (about 4 pages): what ICs need, want and reject, plus what surprised us. It describes what we heard and doesn't argue for a direction. | CTO (Matthew Chan), CPO (Aaron Kurzak), Sunny |
| Day +2 | Shell merge brief: keep / cut / steal for both concepts, plus the Concept 3 ideas | Jeet, Aaron |
| After | The evidence feeds the roadmap discussion. Bugs go to the tracker. | Leadership, engineering |
06What we need
- Sign-off on the format, and backing to release ICs for a day
- Would one of you open the day? (5 minutes, then leave)
- 30 minutes on Day +2 for the readout
- Knowledge Brain Pipeline Run, and Tested for 15 Participants (Groups B and C)
- Three engineers for the day
- Recruiting based on our profile
- One main room and two breakout rooms
- Food and drinks for about 28
- A fairness check of the session 3 prompts and the future scenarios
Timeline (tight): Fri 25 Sep sign-off and the HR request → Mon 28 assign groups (A and B from the sign-up questionnaire, C from the alpha user list) → Tue 29–Wed 30 Group B sets up → Fri 2 Oct team run-through and account checks → Mon 5 or Tue 6 Oct, the day. Thu 1 Oct is a public holiday.
Workshop kit · Draft, Sep 2026 · No participant names or Crypto.com data appear anywhere in this kit.