I build intelligent systems that connect software, models, and real-world hardware.
| System | What it demonstrates |
|---|---|
| Smart Car | Voice → language model → navigation → MCU → motors. A Raspberry Pi system with an MCU safety layer; obstacle stopping and status telemetry are verified on hardware. |
| Air Quality Forecasting | Data ingestion → Airflow → PostgreSQL → machine learning → dashboard. An end-to-end forecasting platform built for repeatable data and deployment workflows. |
| Agent-Q | Screen context → language model → planning → validation → action. A Windows accessibility overlay that keeps a validation step between an agent and user-facing actions. |
| ESP32 RC Race Car | Controller → ESP-NOW → receiver → control loop → motors. A real-time wireless control system with CRC-checked packets, failsafe behaviour, and yaw-assisted steering. |
These systems are intentionally cross-disciplinary. They combine software engineering with the constraints that matter when code controls devices: latency, unreliable networks, sensor faults, power, and safe failure modes.
| Embedded & Robotics | AI & Software |
|---|---|
| Microcontrollers, real-time control, sensors, wireless links, Linux edge systems, and robotics. | AI systems, data platforms, software architecture, ML workflows, and user-facing applications. |
Languages: C · C++ · Python · TypeScript · JavaScript · C# · SQL · Lua
Embedded: ESP-IDF · Arduino · FreeRTOS · ESP-NOW · TinyUSB · I²C · I²S · SPI · UART
AI and data: LLM systems · PyTorch · XGBoost · scikit-learn · Airflow · PostgreSQL
Software and platform: Docker · Next.js · React · Node.js · .NET
I document what a system does today, what has been verified, and what still needs to be proven.