YC Fall 2026 x Moss: The Zero Latency Builder Sprint

YC Fall 2026 x Moss: The Zero Latency Builder Sprint

Build the next generation of AI applications with Moss and make latency your competitive advantage.

Powered by Moss (YC F25), this builder sprint challenges you to turn ideas from Y Combinator's Fall 2026 Requests for Startups into production-ready AI applications where speed matters.

Moss delivers sub-10ms semantic search without a vector database, giving AI agents instant access to the context they need. Build AI experiences that feel truly real-time, whether they run in the browser, on the edge, on device, or in the cloud.

The challenge is simple: Build something where fast retrieval makes the product meaningfully better.

Why Build with Moss?

• Under 10ms semantic search for real-time AI applications
• No vector database required so you can reduce infrastructure and latency
• Runs anywhere, including the browser, edge, device, or cloud
• Built for AI agents that need fast and reliable context
• Production-tested across more than 10 million real-time voice minutes

You will get access to Moss starter kits, technical resources, and AI-powered support so you can go from idea to working prototype quickly.

What Can You Build?

Real-Time Voice and Conversational AI:

- Build voice agents that need to understand and respond instantly for field workers, healthcare, dispatch, customer support, and more. - Use Moss for: sub 10ms context retrieval, real time knowledge access, and low latency agent interactions.

Multiplayer AI and Collaborative Agents:

- Build AI powered workspaces where humans and agents share context and collaborate on long running tasks. - Use Moss for: fast shared context, session history, semantic retrieval, and multi agent state.

Local First AI and The Small Cloud:

- Build privacy first AI tools, browser copilots, developer utilities, or applications that run close to the user. - Use Moss for: local semantic search, on device retrieval, and lightweight AI experiences without heavy cloud infrastructure.

Agent Reliability, Security and Evaluation:

- Build systems that make AI agents more trustworthy through runtime guardrails, evaluation, and context validation. - Use Moss for: fast retrieval, continuous context evaluation, latency tracing, and real time agent guardrails.

Build Fast. Get Feedback. Ship.

• Start with Moss
• Get a working Moss starter kit and begin building in minutes with support for LiveKit, Next.js, Python, and FastAPI.
• Map Your Architecture
• Submit your architecture early and get feedback on your retrieval pipeline, agent state, and potential bottlenecks.
• Build Your Product
• Use Moss to power your retrieval layer and focus on the experience, intelligence, and product you are actually building.
• Ship It
• Submit your GitHub repository, architecture diagram, and a two minute demo showing what you built.

Timeline:

• August 24, 2026 – September 20, 2026: Registration
• Read the complete hackathon details, understand the challenge and requirements, join the official WhatsApp community for updates.
• September 6 – September 20, 2026: Kickoff, Build Phase & Project Submission
• Build your AI application using Moss as the retrieval layer. Submit your Architecture Diagram, PRD, GitHub Repository, Deployed Link of the Agent, and Video Demo by 11:59 PM IST on September 20.
• September 23, 2026: Finalists Announced
• Submitted projects are evaluated based on Product & User Experience, Technical Execution, Speed & Latency, and Demo & Presentation. Top-performing projects will be announced as finalists.
• September 26, 2026: Grand Finale
• Finalists demo their products live before the judging panel at OJone, Bengaluru, followed by the final evaluation, winner announcement, and awards ceremony.

Prizes:

• ₹50,000 in cash prizes across the Top 5 teams
• ₹5,000 community pool for the Top 10 Social Media Builders
• ₹3,000 pool for the Top 20 Referrers
• Partner swag and cloud credits
• Curated technical books
• Digital certificates for qualifying submissions
• Teams can compete individually or in pairs.

How We Will Judge:

• 35% Product and User Experience
• Does the product solve a real problem and is the experience compelling?
• 30% Technical Execution
• How well does the architecture and AI pipeline work?
• 20% Speed and Latency
• Did you use Moss effectively to build a fast and responsive experience?
• 15% Demo and Presentation
• Can you clearly show what you built and why it matters?

Meet Moss:

Moss is the retrieval layer for production AI systems.

When your AI application needs context instantly, Moss delivers semantic search in under 10ms without requiring a traditional vector database.

That means you can build AI systems that respond faster, run closer to users, and avoid unnecessary infrastructure. From real-time voice agents to local-first applications and autonomous AI systems.

Now it is your turn.

Take a YC-inspired startup idea. Build it with Moss. Push the limits of what is possible when AI retrieval is no longer the bottleneck.

Build fast. Retrieve instantly. Ship something real.