Wendy provides an operating system, WendyOS, that lets developers deploy containerized AI applications to edge devices such as NVIDIA Jetson, Thor, and Raspberry Pi with a single command. It supports USB‑C, wireless, and cloud deployment, works offline, and includes fine‑grained security entitlements for GPU, camera, audio, networking, and storage, enabling rapid, secure rollout of physical AI workloads.
Funding
Funding not disclosed
Founders
Product
Problem
Deploying AI applications to edge devices such as robots, drones, and sensors typically requires manual flashing, complex toolchains, and device‑specific setup, which slows development cycles and limits rapid iteration.
Solution
Wendy provides an out‑of‑the‑box operating system, WendyOS, that lets developers build containerized AI apps in their preferred language and deploy them to supported edge hardware with a single command. The system works over USB‑C, wireless, or cloud connections, enabling deployment even when the device has no network access. It abstracts hardware differences across NVIDIA Jetson, Thor, and Raspberry Pi platforms, handling driver integration and over‑the‑air updates automatically. Fine‑grained entitlements enforce security by restricting GPU, camera, audio, networking, and storage access per app. The open‑source stack runs on‑device and integrates with existing development environments on macOS, Windows, Ubuntu, Fedora, and Arch.
Target Audience
Primary users are robotics, drone, and sensor development teams that need fast, secure deployment of AI workloads to edge hardware, as well as engineers building physical AI products across multiple platforms.
Features
- One‑command deployment of containerized AI applications via the Wendy CLI, eliminating flashing and toolchain configuration
- Supports deployment over USB‑C host mode, Wi‑Fi, or cloud, allowing work without internet connectivity on the edge device
- Cross‑platform compatibility with NVIDIA Jetson, Thor, and Raspberry Pi hardware
- Multi‑language support including Python, Rust, Swift, C++, and TypeScript for flexible development
- Built‑in fine‑grained entitlements that sandbox GPU, camera, audio, networking, and storage access per application
- Open‑source Apache 2.0 license with full source availability on GitHub, preventing vendor lock‑in
- Integrated over‑the‑air update mechanism for continuous delivery and iteration on deployed devices