Three takeaways from AWS Summit 2026
Success demands your own stack. AI can build it. The time to start is now.
I was at AWS Summit this morning, an event livestreamed across ASEAN. My takeaway? Success demands your own stack, AI can build it, and the time to start is now.
Success calls for your own stack
What would you do if your firm is on the verge of an IPO, and what you have to make existential decisions on across eight countries is a 47-tab spreadsheet?
That was the challenge faced by Grab's Ken Lek and his team in 2021. So they set out to fix the foundation before chasing the future.
They built a centralised data lake on S3 with open table formats, unifying dozens of SaaS and analytics systems into a single source of truth. This dropped manual reconciliation by 60% while month-end reporting went real-time. Today, the team is deploying AI agents on top to enable natural-language queries and more.
A new way to AI code
What Grab took years to build, AWS now wants every developer to ship in days with AI.
AWS today announced that it is bumping up the number of free credits for its Kiro agentic development environment from the usual 50 free credits to 1,000 for students and adult learners from all polytechnics, ITE, and universities.
I've not used Kiro before, but AWS claims it works differently by requiring a specification-driven approach. Scope, scenarios, and success must first be defined in natural language.
My personal experience with the current vibe-style approach is that while it works, it cannot scale to multi-user deployments. More complex apps will also start to fail. Kiro seems positioned to meet this gap.
Time to start is now
Finally, SMS Desmond Tan reminded us that AI may feel unprecedented, but it isn't the first major disruption Singapore has faced, and probably won't be the last.
He shared the path so far: computerisation in the 1980s, the internet in the 1990s, "Infocomm" in the 2000s, and Smart Nation in the 2010s. So now we have generative AI.
His point? Each previous wave brought similar anxieties around jobs, skills, and potential widening of inequality. Yet each time, Singapore emerged stronger, because we didn't sit it out.
The companies and individuals who start now, even imperfectly, will compound their learning while others are still debating whether to begin.
Start small. Start messy. But start today.
