Independent AI implementation for Hong Kong leadership teams. I find where coordination breaks between individual productivity and organizational results, build a working version of an AI-first way to coordinate, and hand off the playbook so your team can keep going.
Most companies have done the easy parts. Bought the tools, trained the staff, run the pilots. Individuals got faster. The organization didn’t follow.
of CEOs report no financial benefit from AI investment so far.
say it has delivered both cost savings and revenue gains.
The reason isn’t tools or training. It’s coordination, and nobody has done the unglamorous work of rebuilding how the organization runs around faster people.
PwC 29th Global CEO Survey · January 2026
Map where work actually breaks between teams. Find the decisions that get made, lost, and re-litigated. Surface what the data shows that internal reports don’t.
A working version of an AI-first way to coordinate, running on your real work. Custom Claude Skills, agents, automations. Tools your team uses, not slides they file.
Documentation and capability transfer so your team extends the work after I leave. No dependency.
The systems run on trusted platforms your organization controls (Claude, MCP integrations, vendor tools). For anything needing production hardening at scale, I produce specs your engineering team or partners can build against.
A leadership team had rolled out AI across every department. Output went up, but the weekly leadership meeting was still two hours of status synchronization. Department heads spent the night before manually pulling together fragmented reports, summaries, and updates from across the organization.
We redesigned how reporting, decision prep, and context moved through the week, then built the systems that kept synchronization running continuously in the background. The coordination work that used to happen manually the night before now happened continuously across the organization.
The leadership meeting stopped being a status update. Leaders walked in already synchronized. The meeting became the place where actual decisions got made.
A representative example of the coordination problems Unlearn Labs works on.
Two ways to work together, from a single high-drag workflow to your whole operating model. We scope the right fit in conversation.
One high-drag workflow rebuilt and the systems to support it, plus a methodology playbook your team keeps.
6–8 weeks · One workflow
Ongoing coordination work across multiple workflows, with capability transfer to your team.
Ongoing · Multi-workflow
Engagements scale from mid-sized teams to enterprise. Eligible Hong Kong enterprises can apply Technology Voucher Programme funding (up to 75% subsidy).
Six short questions. One personalized read on where AI is reshaping your role and your institution. Built for leaders and operators who run real work, not AI demos.
Take the Compass →Product management at Bybit (GameFi), Animoca Brands, and Zynga. Built and operated loyalty systems used by enterprise partners across multiple sectors.
Founded and ran Allegianc3, a Web3 loyalty platform recognized at the 2024 Hong Kong Business Awards. Running and eventually winding down a venture taught me more about coordination breakdown than any consulting framework could.
Now applying that to one problem: closing the gap between AI investment and organizational return for Hong Kong companies. Works directly with leadership teams. No account managers, no offshore delivery teams.