Albert Au 區恩庭

Writing

Albert Au writes in three series: HealthTech, Social Innovation, and Lean LaunchPad and Entrepreneurship.

HealthTech

AI can fold proteins and draft lab plans — clinical development still decides who wins

2 Nov 2026, on LinkedIn. Folding a protein or drafting a lab plan is not the finish: clinical development, mechanism, and translation still decide which drug-discovery story reaches patients.

Accurate AI still gets abandoned — SHAKED’s mid-shift lesson for Asia operators

26 Oct 2026, on LinkedIn. In Nature Medicine’s SHAKED emergency-department pilot, 99 of 100 sampled outputs were clinically appropriate, and clinician use still fell from about 68% to about 30% over four weeks.

Symptom apps only work when a nurse answers — Hong Kong’s SUPPORT+ proof

19 Oct 2026, on LinkedIn. HKUMed’s SUPPORT+ cut unplanned hospitalisations in a 1,214-patient palliative-care trial because a nurse answers when symptoms turn severe.

Is AI ROI the wrong metric? APAC is asking about workflow fit

12 Oct 2026, on LinkedIn. Part 2 after the 5 Oct essay: HIMSS26 APAC and a survey of 200 APAC providers on whether workflow fit, not classic ROI, decides what scales.

AI won't replace the doctor. Poor implementation will hold it back first.

5 Oct 2026, on LinkedIn. What Albert Au took from Professor Tien Yin Wong's keynote at the CHART inauguration, HKU Faculty of Medicine.

Asia HealthTech: registration is a living vigilance system — not a one-time PDF

28 Sep 2026, on LinkedIn. Singapore's HSA GN-15 R14 and Taiwan's AI regulatory tools, and what operators shipping SaMD and Class C medical apps should keep watching.

In Asia HealthTech, the regulatory pathway is the product

21 Sep 2026, on LinkedIn. A fortnight of clearances from Singapore, China, and Japan, and what founders should underwrite before they scale.

Social Innovation

Social Innovation Anywhere — Call for Fringe Event Partner

28 Oct 2026, on LinkedIn. A call for partners to host social-innovation events across Hong Kong from 1 to 18 December 2026, outside the Social Enterprise Summit main room. Confirm by 6 November 2026.

What SES 2026 is teaching me as deputy chair

21 Oct 2026, on LinkedIn. What the deputy-chair seat is teaching him: cross-sector listening, a wider speaker search, Impact Talk, and the coordination gaps he is still closing.

Social Innovation Anywhere (看見需要・共創幸福)

13 Oct 2026, on LinkedIn. How the Social Enterprise Summit 2026 organising committee framed this year’s theme, starting the exchange with Macau on 10–12 December.

Social enterprise is the form. Social innovation is the method

7 Oct 2026, on LinkedIn. What fifteen radio conversations with Hong Kong social innovators taught him, and why he is publishing them again on HKGoodPost. The Chinese essay is 社企是形式，社創是方法｜區恩庭：同理心要驗證，驗證之後要營運.

In Hong Kong social innovation, community discovery is the product — not the pitch deck

30 Sep 2026, on LinkedIn. Also on HKGoodPost, 4 Oct 2026. What the SVhk–Sino Impactathon reminds social entrepreneurs, NGOs and funders about learning with communities before they scale.

Digital inclusion starts with confidence, not features

23 Sep 2026, on LinkedIn. Also on HKGoodPost, 30 Sep 2026.

Albert Au also writes on HKGoodPost (好報), the SDG and social-innovation platform he co-founded.

Lean LaunchPad and Entrepreneurship

Capture VPC from interviews, not the whiteboard

29 Oct 2026, on LinkedIn. How to fill the Value Proposition Canvas from interviews, observation, and secondary research, so every sticky has a source.

Why VPC finds real customer value

22 Oct 2026, on LinkedIn. Why the canvas separates the customer’s world from the offer, and treats fit as a claim until customers confirm it.

Double bottom line is not two jobs

15 Oct 2026, on LinkedIn. An impact metric that lives only in the pitch appendix is not a double bottom line: customer, impact, and money have to sit on one Social Lean Canvas.

Bring me the sentence that surprised you

8 Oct 2026, on LinkedIn. The homework he now sets before a coaching session: one sentence a customer said that the founder did not expect.

Self-assess in order: TRL, then CRL, then IRL

1 Oct 2026, on LinkedIn. Why founders score technology, commercial, and investment readiness in that order. The check is on this site: Founder readiness check.

Your MVP looks finished. Your customer discovery isn't.

24 Sep 2026. First published on LinkedIn. What Stanford Lean LaunchPad 2026 is testing, and what mentor rooms in Hong Kong keep seeing.