Embedded Software Engineer – AI/ML at the Edge | Hybrid – Berlin, Germany
Are you passionate about pushing intelligence to the very edge of devices? Do you thrive in the intersection of AI/ML, embedded systems, and extreme hardware constraints?
We’re looking for an Embedded Software Engineer with deep experience in Edge AI and TinyML to help us build our client's smart, low-power, real-time solutions for next-generation IoT, wearables, and embedded vision systems .
This is a hybrid role based in Berlin, where innovation in edge computing, machine learning, and embedded tech is booming.
What You'll Do:
* Deploy and optimize ML models on ultra-constrained edge devices
* Work with TensorFlow Lite for Microcontrollers, Edge Impulse, or Apache TVM
* Develop real-time embedded applications in C/C++, integrating sensor data pipelines and ML inference
* Tune performance, latency, and memory footprint for MCUs (ARM Cortex-M, RISC-V, etc.)
* Collaborate with data scientists to convert and quantize models for on-device execution
* Design robust OTA mechanisms, fault handling, and power management strategies
What You Bring:
* Experience deploying TinyML models in production or prototypes
* Proficiency in embedded C/C++, RTOS-based development, and low-level debugging
* Hands-on with MCUs, including memory constraints, power optimization, and peripherals (I2C, SPI, ADC, BLE)
* Familiarity with TensorFlow Lite for Microcontrollers, TVM, or Edge Impulse
* Ability to perform model quantization, pruning, and conversion for low-footprint deployment
* Experience with Python for model training, tooling, and integration scripting
Nice-to-Have:
* Experience with hardware accelerators (e.G., Ethos-U, NPU, DSPs)
* Knowledge of micropower design and battery-operated devices
* Familiarity with OTA updates, secure boot, or embedded Linux environments
* Exposure to real-time sensor fusion, anomaly detection, or embedded vision use cases