Agentic AI Build Week 2026: The Largest Buildathon ASEAN Has Ever Seen

On July 12, five days of 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗔𝗜 𝗕𝘂𝗶𝗹𝗱 𝗪𝗲𝗲𝗸 (AABW) closed with the largest buildathon ASEAN has ever held.

This wasn’t the first swing GenAI Fund took at building Southeast Asia’s AI builder ecosystem this year. It was the third, and the biggest one yet.

The numbers:

  • 🧑‍💻 3,000+ builders across 55 countries, from Southeast Asia, India, the US, and Europe 
  • 🚀 400+ projects went in front of judges
  • ⚖️ 70+ judges evaluated every one of them across two full rounds of pitching

The Road to AABW

In March, LotusHacks brought over 1,500 builders from 30 countries to VNG Campus for 36 hours straight, the largest AI hackathon Vietnam had ever seen. GenAI Fund came in as co-organizer, backing a group of highschoolers with a vision nobody had attempted at that scale.

In April, Qwen AI Build Day moved the format forward. GenAI Fund and Alibaba Cloud put 300+ builders directly in front of enterprise problems across Mobility, Financial Services, Healthcare, and Public Sector, judged by the people who actually own those problems inside Tasco, Shinhan Future’s Lab, Elfie, and Logivan.

Each event proved a different piece of the thesis. LotusHacks proved the talent pool was real. Qwen AI Build Day proved builders could solve enterprise problems, not just demo projects. Agentic AI Build Week was built to prove both at once, at a scale no one in the region had tried before.

What AABW brought to the table:

  • 🧑‍💻 3,000+ builders across the week
  • 🏢 11 tracks, 9 enterprise-led plus 2 track problems run by GenAI Fund itself
  • 🎮🤖 Southeast Asia’s first Gaming AI and Robotics AI hackathon categories
  • 💳 $2M in claimable AI credits across the partner stack, including OpenAI, Kimi (Moonshot AI), Agora, TinyFish, Bright Data, Notion, and TRAE
  • 💰 A $2M prize pool, including proof-of-concept budgets and hiring pipelines straight from the enterprises that own each track’s problem

July 8–10: Three Days of Workshops

Before a single line of hackathon code got written, three straight days of workshops ran across three venues: Tasco office, the AWS office in Bitexco Financial Tower, and VNG Campus.

Every single day sold out. Every single day ran packed and full, 300+ builders in the room from open to close, led by the teams behind the platforms builders would actually use to build.

📍 Day 1 · Tasco Office · 5 sessions, officially full, standing room only

  • 🎨 BytePlus — Render the Next Era of Creation with the BytePlus AI Stack — Bui Kien, Solution Architect Lead & Tran Quoc Huy, Enterprise Sales Manager
  • ☁️ Tencent Cloud — The Full-Stack Advantage: Building Production-Ready AI Agents — Max Meiden Dasuki, Dex Guo, Frida Feng & Kai Zeng, Solutions & Product Architecture team
  • 🚀 NVIDIA — Inside NVIDIA Inception Program: How Startups Build & Scale AI Globally — Jason Pham, NVIDIA Inception Community Manager SEA, with guest founder Trieu Vu Duy, CTO of NYB.AI
  • ⚡ TRAE — TRAE in Your Professional Workflow — Max Wan, Product Operations Manager & Rita Huang, Global Community Growth
  • 🤖 OpenClaw — From Personal Automation to Business Workflows — Goon Nguyen, Founder of GoClaw.sh & CTO of TOP Group, and Ky Nam, Founder & CEO of Flowser.ai

📍 Day 2 · AWS Office, Bitexco Financial Tower · 5 sessions, 100%+ capacity from the morning open

  • 🧑‍💻 AWS — From Spec to Production Code: Kiro, Claude Code & Codex on AWS — Diep Nguyen, Solutions Architect
  • 🦾 Agora — Physical AI Party: Build Voice Agents with Agora ConvoAI — Derek Zheng, Head of Developer Relations APAC. Builders assembled and coded physical hardware kits live in the room. One team that touched hardware for the first time in that session went on to win the Robotics & Physical AI track four days later.
  • 🧠 AWS — Production Multi-Agent AI on AWS, Bedrock AgentCore — Thi Nguyen, Senior Solutions Architect
  • 🐟 TinyFish — Enable Track Live Demo Workshop — led live by the TinyFish team
  • 🛠️ AWS — Design Patterns and Best Practices: Testing, Monitoring and Production Readiness — led by the AWS Solutions Architecture team

📍 Day 3 · VNG Campus · 6 sessions, room still packed by the last one

  • 🕷️ Apify — Build, Deploy & Monetize AI Agents — Jakub Drobnik, Engineering Team Lead & Saurav Jain, Senior Developer Community Manager
  • 🔍 Langfuse — Enterprise AI Track: LLM Observability & Evals with Langfuse — led by the Langfuse team
  • 🧑‍🏫 Google Developers — Beyond Autocomplete — led by the Google Developer Experts team
  • 🔐 Antitech — Securing Agentic AI — led by the Antitech team
  • 💸 Virtuals Protocol — Autonomous Internet Payments via EconomyOS — Celeste Ang, Head of DevRel & Joey, DevRel
  • 🪪 Terminal3 — Setting Up Enterprise-Ready Agent Identities to Qualify For Regulated Workflows — Ian Chong, Devrel Lead

By the time Day 3 wrapped, builders had three full days of hands-on technical grounding before a single hackathon problem statement was even locked in.

🎬 July 11: Kick-Off Day

Three months of planning. Three workshop days already in the books. Today, the hackathon itself officially began.

𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗔𝗜 𝗕𝘂𝗶𝗹𝗱 𝗪𝗲𝗲𝗸 kicked off at Galaxy Innovation Hub (GIHUB), the venue where every build, every judge session, and every pitch would happen through Demo Day. Laura Nguyen opened the morning, welcoming builders into the space.

Three keynotes set the tone for what was ahead:

  • 🎤 Tuoc Huynh, CEO, Galaxy Holdings
  • 🎤 Dr. Benedict The, Chief Digital & Technology Officer, VietJet Air
  • 🎤 Girish Dilip Patil, Tech Lead for Applied AI, AWS

Three different worlds. One shared message: the opportunity is abundant, the domain expertise in the room is real, and what gets built in San Francisco can just as well get built in Ho Chi Minh City.

Two more sessions ran that morning, led by the GenAI Fund team itself:

  • 📋 Best Practices for Hackathon Submissions — Eugene Koon, Head of Programs, GenAI Fund
  • 🎤 Best Practices for Pitching — Kai Yong, Partner, GenAI Fund

Eugene walked builders through the dos and don’ts of submitting on the AABW Portal, how to lock in their track, and what actually makes a submission stand out to a judge. Kai Yong broke down what a pitch needs to land in seven minutes flat, the same clock every team would be racing against on Demo Day. Builders who sat in that room went into the overnight build already knowing what the people evaluating their work were actually looking for, insider context most teams at a hackathon this size never get.

By 3PM, every team had locked in their track, in person, as required. By nightfall, the overnight build was underway. Builders who signed up for the late night stretch had their sleeping bags out for the 24 hours between Kick-Off and the Demo Day submission deadline.

What followed was pure momentum. Energy drinks, code that refused to cooperate at 3AM, builders scattered across every corner of GIHUB. Some teams still caught the World Cup at 4AM before jumping right back into their builds. The youngest builder in the room was 13, pulling an all-nighter with a smile that impressed everyone around them.

🏆 July 12: Demo Day

Demo Day filled the Main Hall at GIHUB.

400+ projects went through two full rounds of pitching in front of 70+ judges spread across multiple stations, evaluated on six criteria: agentic AI use, problem and track fit, technical execution, impact and usefulness, creativity, and pitch performance.

Round one narrowed each track to its top teams. A 90 minute gap gave the shortlist time to tighten their demos before round two, where they walked track owners through the build in depth instead of just repeating the pitch.


By the end of the day, 11 tracks each had the winners.

🥇 Track Winners

GenAI Fund-led Track Winners

  • 🚀 Founder Track  → WeMaidAi & BrainSquared
  • 🤖 Robotics & Physical AI → Elemental Arena

These two teams aren’t walking away with a trophy for the weekend. They’re walking away with a potential check from GenAI Fund and its network of investors to go build the company.

Enterprise Track Winners (9 tracks, 10 real enterprise problems)

  • 🚗 Mobility (Tasco) → Ask Windy Everything, RouteMate, Car Pooling
  • 🍔 F&B (KFC Vietnam) → Twohearts
  • ✈️ Aviation (Galaxy Holdings & VietJet Air) → Aerotrace Tire Intelligence
  • ☁️ Built with AWS → One Team
  • 🛍️ Retail (Phong Vu) → Intelliworks
  • 🛡️ Retail Guardian (Guardian Vietnam) → Guardians of the Galaxy
  • 🏨 Hospitality (The Anam) → AltoTech
  • 🎮 Gaming (VNGGames) → Agamotto
  • 💰 Financial Services I (Shinhan Future’s Lab Vietnam) → Little Boy’s Aegis
  • 💳 Financial Services II (GoTyme) → Botanary

Thirteen teams. Ten real problems. Every one of them was built and pitched in the same week.

Thank You

None of this happens without the partners who put real problems, real credits, and real people in the room.

🏢 Enterprise Track Partners, who brought their actual business problems and budgets to the table:

AWS, Tasco, KFC, VNGGames, Phong Vu, Guardian, Galaxy Holdings, VietJet Air, Shinhan Future’s Lab, GoTyme, and The Anam.

🛠️ Tech & Ecosystem Partners, who powered the credits, tools, and infrastructure builders ran on all week:

AWS, OpenAI, NVIDIA, Tencent Cloud, Microsoft for Startups, Google for Developers, Qwen, Kimi (Moonshot AI), Netlify, Qoder, Agora, Langfuse, ClickHouse, TRAE, Notion, ZenRows, BytePlus, Bright Data, Apify, TinyFish, Antitech, n8n, Daytona, Featherless AI, Virtuals Protocol, Terminal3, Datadog, and Crustdata (YC F24).

And to every builder who showed up, stayed up, and shipped: you’re exactly why we do this. 🔥

Keep Building With Us

If this recap made you want the full picture of who is building AI across Southeast Asia right now, the State of AI Builders in Southeast Asia 2026 Report has it.

📥 Download the free preview, or unlock the full report 👉 https://aitalent.genaifund.ai/report-2026

Every purchase funds AI access for underserved builder talent across the region.

This Is Only the Beginning

Three thousand builders showed up this year. Some had never shipped production code before Day 1. By Demo Day, they were pitching to the people who could fund what they build next.

That’s the real story of Agentic AI Build Week. Not the credits, not the prize pool, not even the deals on the table. It’s proof that a region can stop asking for a seat at the table and build its own instead.

The talent is here. The energy is real. And this is just the first chapter.

See you at the next one. 👋