One person. 14 AI agents.

This is how we build complex embedded systems.

A reference project: wireless IoT device with hybrid C/C++/Rust firmware, custom hardware, and hardware-in-the-loop testing.

14
AI Agents
11
Repositories
66+
Hardware Tests
26k+
Data Points

The Agent Ecosystem

Each box is a specialized AI agent with deep knowledge of its domain. New agents go through a structured onboarding — learning the codebase, registering with the team, and building domain expertise before contributing.

DEVELOPMENT Firmware C++ / C / Rust / Zephyr Desktop App Rust / Iced CLI Tool Rust / BLE Shared Library Protocols / Types HW Design PCB / Mechanical Edge AI PyTorch / MCU Marketing Website / Campaigns TESTING Test Automation Python / BLE / Power Test Hardware FW Wired to device under test DK Integration Simulated peripherals CI / Nightly Tests Automated on every push OPERATIONS Documentation 150+ docs, continuously updated Infrastructure VMs / CI / Databases Market Research 7 sources / 4h polling Supply Chain ERP / Inventory GitHub Issues · Pull Requests · CI/CD Shared Status File Read & written by all agents

How They Collaborate

Multiple agents contribute insights and requirements Research Testing HW Design Docs Supply Chain GitHub Issues & PRs Agent Implements Code, design, content — in its own repo Validated on Real Hardware Automated tests, power profiling, BLE verification Feedback loop Documented & synced to all agents
||

Parallel by default

Firmware, testing, hardware design, market research, and infrastructure all advance in the same cycle. No waiting.

>_

Deep domain knowledge

Each agent carries full context of its repository — build systems, hardware interfaces, conventions, and architectural decisions.

Hardware-validated

66+ automated tests on real hardware. Power profiled on every change. Not simulations — measurements.