AI hardware and IoT devices
Circuit boards, firmware, edge AI and the cloud behind them. We build connected devices that sense, decide and act, from chip to cloud.
- Node 14 needs waterSoil moisture under 18% in zone BOpen
- P3Pump 3 scheduled05:40 for 25 minutes
- Firmware 2.4 rolled out48 of 48 nodes updated
The problem we solve
Sensors produce data all day, but data on its own changes nothing. Devices that only report leave someone to read the numbers and decide what to do.
We design the board, the firmware and the cloud together, and put AI where it belongs: on the device when it must react in milliseconds, in the cloud when it needs the whole fleet’s history. Our own product, Shaduf, is built this way.
Typical engagement
- Feasibility (2–4 weeks): requirements, a component plan and a working proof on development boards.
- Prototype (8–16 weeks): custom boards, firmware, edge models and the cloud platform, tested in the field.
- Production: design for manufacture, over-the-air updates and fleet monitoring.
What we deliver
- Requirements, component choice and power budget
- Schematic capture, PCB layout and prototype bring-up
- Firmware with on-device AI where the cloud is too slow or too far
- Connectivity over Wi-Fi, BLE, LoRaWAN, cellular or CAN bus
- A cloud platform, agents and dashboards for the device fleet
Tools we use
We pick tools to fit your systems and your team, and these are the ones we reach for most.
KiCad
C and C++
Rust
ESP32
- STM32
TensorFlow Lite
- MQTT
Related work
- CAN bus configuration tool
A cross-platform desktop tool for configuring vehicle control modules, with Dart FFI bindings to the C++ core and CAN bus simulators for testing.