Latent AI: Inference at the Edge

WebAssembly AI services to process data closest to the source and perform inferences at the distributed edge.


Latent AI accelerates AI implementation and workflows for the enterprise, rapidly and cost-effectively anywhere on the edge continuum. We provide deep learning and AI developers with game-changing software tools that enable them to compress, compile and adapt AI models to any hardware target, OS, framework or architecture, faster and more accurately than ever before. Latent AI’s SDK is based on DARPA technologies and is an award-winning spinout from SRI International.

Why Workers?

One reason we built our application on Workers is that it provides scalability on the distributed Cloudflare network—it is simply impractical to send all data to a centralized cloud. Another reason is performance: when AI services are distributed away from centralized data centers and closer to the service edge, it becomes possible to enhance the overall application speed without moving data unnecessarily. Lastly, Workers already extends support for WASM through their Wrangler CLI tool, which is designed to enable a smooth developer experience.