A two-round APAC builder challenge. Build a live AI Employee on Supervity Auto.
Sprint online. Finish on stage in Kuala Lumpur.
Your challenge: Build a live AI Employee on the Supervity Auto platform, one that executes real work, not generated text.
It connects to real systems and acts on them
It follows enforced AI Policies
It handles exceptions gracefully through Auto Workbench
No generated-text-only submissions
No mock demos or slideware
Every Round 1 build must ship with all five:
1 Operator Agent solving a scenario from one of the five tracks
3+ Real Integrations- Microsoft 365, Slack, REST APIs, GitHub, Supabase, or similar
1+ Dynamic AI Policy enforced in the build
1+ Exception handled through Auto Workbench
Real execution against a system, not generated text
Submit with a 3–5 min video walkthrough, your Auto workspace link, a one-paragraph business case, and your track tag.
July 18–19, 2026 · Online across APAC Description
A 48-hour online sprint. Pick a track, build on Supervity Auto, and submit a working AI Employee
Mandatory live platform workshop · July 11–12
Scenario pack with a real company, persona & data
Evaluation July 20–23 · Finalists announced July 25
Aug 3–9, 2026 · Remote build → Kuala Lumpur
Finalists evolve their Operator Agent into a full Auto App and demo it live at Asia Pacific University.
Multi-agent orchestration: Orchestrator + Operator pattern
Custom front-end via GitHub starter repo- NextJS, FastAPI, Docker
10-minute live showcase to enterprise judges · Aug 9
Every track ships with a scenario pack, a fictional company, a persona, sample data, and success metrics. You're solving a brief, not building in a vacuum.
$5,000
on the table paid out in every one of the five tracks