AI Singapore’s Vision for the Region
AI Singapore (AISG) is a national AI programme based in Singapore, focused on building real, homegrown AI capabilities for the region. Its flagship model, SEA-Lion, is a large language model built specifically for Southeast Asia – multilingual, regionally grounded, and designed to reflect the diversity of languages and cultures across the region.
The PAN SEA-Lion Developer Challenge set out to do three things:
🎯 Advance multilingual model building: Helping teams build AI applications that serve speakers of Southeast Asia’s many languages, not just English.
🎯 Accelerate market-ready AI innovation: Pushing solutions beyond experimentation toward real commercial viability, with winners pitching at SWITCH.
🎯 Grow SEA-Lion adoption: Getting developers to build directly on SEA-Lion and building a community around it for the long term.
AngelHack as the Engine Behind the Challenge
AngelHack was trusted by AISG to plan and run the challenge from start to finish – a four-week virtual competition with finalist pitches at SWITCH. AWS and Google came on board to provide cloud credits, while IMDA and NVIDIA brought in expert speakers to share what AI adoption actually looks like on the ground across Southeast Asia.
Before the virtual challenge kicked off, we ran an in-person event at ATxSingapore to get things moving. The session introduced attendees to SEA-Lion, walked through use cases where it had already shown real results, and gave people a reason to sign up. The idea was simple: show that SEA-Lion was ready, then invite people to build with it.
The challenge itself was organised around four verticals – Healthcare, Finance, Education, and Public Sector. Teams picked one and defined a specific, locally relevant problem that SEA-Lion could help solve. Throughout the four weeks, workshops gave teams the regional context they needed to ideate well, and mentorship sessions helped them pressure-test ideas and make sure solutions stay grounded in real need.
Participation and Impact
From first sign-up to final pitch, participation stayed strong at every stage.
- 796 total registrations
- 236 Round 1 submissions (AI judging)
- 40 Round 2 finalists (panel judging)
The numbers tell part of the story, but what stands out is the quality of engagement throughout. Teams weren’t just signing up – they were building real solutions for real problems in healthcare, finance, education, and the public sector across Southeast Asia. By the end, SEA-Lion had something arguably more valuable than metrics: a pipeline of developers who know the model, have shipped something with it, and are ready to go further.
