Beta Fund X AWS X OpenAI Multi-Model Hackathon
🛑 AI is moving beyond text.
The next generation of products will reason across images, video, audio, voice, documents, and real-world data—not just prompts and responses. This hackathon is about building for that future.
Join us at AWS Builder Loft on October 23 for a one-day Multi-Modal Hackathon, bringing together builders, researchers, founders, and investors to ship products that combine multiple forms of intelligence into something people can actually use.
This is a build-first event. Come with an idea, find a teammate, and leave with a working demo.
👉 Register here for AWS Builders Loft Entrance: https://events.builder.aws.com/W4PEkn
Loft Policy
All guests must be 18+ and present a valid government-issued physical photo ID at check-in (digital IDs will not be accepted). Scooters and bikes are not permitted in any Amazon building; no on-site parking available.
Format
2–4 builders per team
One-day hackathon
Working demo required
Technical credits and resources will be provided
Additional sponsor and integration requirements will be announced
📅 Schedule
9:30 AM — Check-in & Doors Open
10:00 AM — Opening Remarks + Sponsor Workshop
10:30 AM — Panel / Speaker Session (TBA)
11:15 AM — Build Starts
11:30 AM — Lunch (provided by Keenable)
1:00 PM — Build Continues
5:00 PM — Demo Showcase
6:30 PM — 🏆 Awards & Closing
Focus Tracks
👁️ Multi-Modal Applications
Build products that combine multiple forms of input—such as text, image, video, voice, audio, or documents—to solve real user problems.
🤖 Multi-Modal Agents
Build agents that can see, hear, reason, and take action across different modalities, from screen-based workflows to voice, vision, and document understanding.
🧱 Multi-Modal Infrastructure
Build the underlying tools that make multi-modal systems work, including retrieval, memory, evaluation, model routing, data pipelines, observability, and developer infrastructure.
Full track details and judging criteria will be announced soon.
Prizes & Opportunities
🏆 Cash Prizes
1st: $500
2nd: $300
3rd: $200
🔎 Keenable Prize
$1,000+ in Keenable credits for the winning team, judged by George Trushevskiy (Keenable).
🧪 Open Challenge: MatrAIx Persona 1M
Multi-modal models still break in ways that are hard to see until real users hit them. We're inviting builders to help find where.
Run MatrAIx Persona 1M on image, audio, or video generation models
Submit failure cases at personalagenteval.com/contribute
Keep contributing after the event → become a co-author on our Personal Agent Eval project!
Prize pool and placements to be announced.
📹 Best Use of FaceTime (Chert)
$300 + 1 month of free dedicated Chert line + guaranteed Software Engineering interview at Chert (YC P26), judged by Gary Gao (Chert).
AI agents can now see, show, and proactively call users on FaceTime. Show us what becomes possible when your agent can pick up the phone and look you in the eye.
Build on Chert's FaceTime API (add iMessage if you want)
Join, submit, and showcase at the Chert Discord: Join Discord Server
Submit: a repo, a LinkedIn/X post, and a demo video of a live FaceTime call (under 2 minutes)
💰 Beta Fund Fellowship
Standout founders and teams may be invited to interview for Beta Fund's Fellowship and potential investment.
Partners
Chert lets anyone put a realtime AI video agent on a live FaceTime call with one API. We're already powering teams building AI companions, coaches, real-estate, and onboarding agents.
Keenable (keenable.ai), the independent web search API for AI — Integration Partner. Every onsite builder gets $200 in Keenable credits (details shared at check-in). Keenable is also providing lunch.
MatrAIx (matraix.ai) — Open Challenge Partner.
Venue
📍 AWS Builder Loft — San Francisco
📅 October 23, 2026
More details on sponsors, tracks, judges, speakers, prizes, and technical requirements coming soon.
For now: build something that can see, hear, understand, and act.