We’re building the commerce platform for the places traditional POS can’t reach.

The default assumption of the last twenty years of commerce software has been that the cloud is always there. It isn’t. Festivals, ships, food trucks, remote sites, disaster zones, and restaurants under network stress all live with that gap every day, and they pay for it in lost revenue and bad decisions.

We exist to close that gap. Not by adding offline mode as a fallback to a cloud-first product, but by inverting the assumption entirely. Local is primary. Cloud is optional. Every device is a peer. Every device is intelligent. The network is no longer a precondition for selling.

Local-first software has already won in collaborative editors, in distributed databases, in design tools, in note-taking, in development environments. Commerce is one of the last categories that still treats the cloud as the source of truth. That’s an artifact of when commerce software was built, not a property of the category itself.

The next ten years are going to flip this. We want Zerobeat to be the platform that flips it. Mesh-native, AI-native, cloud-optional commerce, running on the silicon that’s already in the operator’s hand.

Most of the POS industry treats offline as a feature flag and AI as a cloud integration. We treat both as load-bearing parts of the architecture, and we treat the combination of them as where the actual moat lives.

1. Peer-to-peer mesh, not hub-and-spoke.Every terminal is a full peer. There’s no designated primary device whose failure takes down the venue. Square and Clover ship siloed offline. Toast ships single-primary local sync. We ship a real peer mesh, and the difference is decisive at scale.
2. CRDT-replicated venue state. Conflict-free replicated data types let every terminal hold a full local copy of the relevant operating state, with concurrent edits guaranteed to converge. No coordinator, no last-writer-wins, no manual reconciliation after a partial outage. This is the part that distributed systems engineers will recognize as load-bearing.
3. On-device AI as a first-class dependency.Apple’s Foundation Models, Ollama, MLX, and a maturing quantized open-model ecosystem mean purpose-built models can run on the same iPad that runs the POS. Decisions stay local. Costs stay flat. Data stays at the venue.

The architectural thesis in full lives at Offline-First Is the Future of Live Event POS. If that thesis holds, the addressable market is much bigger than events alone.

We’re leading with festivals because they’re the most punishing commerce environment in the world, and a product that wins there is a product that wins almost everywhere else by definition. But the same local-first, mesh-native, on-device-AI architecture maps cleanly onto a much broader market.

Each of these is a real category with real spend, real pain, and the same underlying architectural failure mode. Building once, deploying broadly, is a rare property of vertical software. We have it because the architecture, not the use case, is the durable thing.

• Live events and high-volume venues are a $100B+ global concessions and ticketing category. The incumbents (Square, Toast, Clover) were built for coffee shops and restaurants and have failed to seriously address it.
• Apple’s Foundation Models framework, shipped in iOS 26, made on-device LLMs a first-class platform feature. Every iPad already has the silicon and the SDK.
• Open-model quantization has matured. Sub-gigabyte purpose-built models running at useful speeds on consumer hardware are a real production option as of 2026, not a research preview.
• CRDT-based local-first architectures are well-understood, with a deep bench of production patterns from collaborative software, distributed databases, and local-first apps.
• Connectivity expectations at events are getting worse, not better. Crowd density is rising. Phone penetration is at saturation. 5G doesn’t solve LAN contention. Starlink doesn’t solve LAN contention. The architecture has to change.

The piece that’s missing isn’t compute or connectivity. It’s a commerce platform built around the way the world actually works at the edge. That’s the platform we’re building, and the window to define this category is now.

• Aligned investors who care about distributed systems, on-device AI, local-first software, and category-defining infrastructure plays, not just SaaS multiples.
• Operators and connectors in live events, venue management, concessions, ticketing, hospitality, and adjacent verticals (cruise, mobile, remote ops) who can open doors to the people who actually run these venues.
• Strategic partners in payments, hardware, connectivity, and AI infrastructure whose roadmaps benefit from a local-first commerce platform becoming the default at the edge.