Experience
Time-series deep learning, signal processing and embedded systems — 2020 to today.
Senior Algorithm Engineer
TDK InvenSense — Grenoble, France
- Designing, training and evaluating production-grade ML for noisy, multi-modal sensor data — classical models alongside CNNs, RNNs/LSTMs and Transformers
- Time-series forecasting, anomaly detection, gesture recognition and spatial tracking on MEMS sensor streams
- Transformer-based architectures adapted to sequential data — time-series Transformers, behavioural embeddings and tokenized sensor streams
- DSP and feature engineering: filtering, windowing and FFT to calibrate raw sensor data and extract robust features
- Edge optimisation for resource-constrained targets — quantization, pruning and hardware-accelerated compilation (ONNX, TensorRT, TensorFlow Lite)
- ML pipelines for high-volume streaming, asynchronous telemetry at low latency; Zephyr RTOS and embedded Linux integration
- Experimental characterisation of microphone, accelerometer, gyroscope, ultrasonic time-of-flight, pressure and magnetometer sensors across environments
Research Engineer — BLE Channel Sounding
Indoor positioning on Nordic nRF54L15 — Nantes, France
- Real-time time-of-flight ranging algorithms (IFFT, MUSIC, ESPRIT) under Zephyr RTOS for BLE 6.0 Channel Sounding
- Hardware configuration, antenna driving, ranging synchronization — production-grade C with code-coverage instrumentation
Doctoral Researcher
GEOLOC Laboratory, Université Gustave Eiffel — Nantes, France
- Physics-informed anomaly detection on multi-modal navigation time series (IMU, GNSS, 5G)
- Autoencoder families (AE, VAE, USAD) for unsupervised detection, and the evaluation methodology behind them
- Embedded ML deployment on resource-constrained microcontrollers
- 10+ peer-reviewed publications in IEEE journals and conferences
- Thesis defended June 2026
Embedded Linux Engineer
Industrial & consumer product projects — Remote, USA
- 10+ production Linux drivers (IIO, serdev, platform) for MEMS and industrial sensors on STM32MP1 and NXP i.MX93
- Full Yocto systems: meta-layers, BSP integration, device-tree bring-up, secure boot, SWUpdate OTA
- Muzziball consumer audio device and truck monitoring platform (CAN, 5G, GNSS, FreeRTOS)
Teaching Assistant
Signal processing, electronics & embedded systems — Aleppo, Syria
- Led 50+ lab sessions — 150+ students trained on instrumentation and sensor measurement methodology
Education
PhD, Signal Processing & Navigation
Université Gustave Eiffel, Nantes, France
Thesis: Physics-Informed Anomaly Detection for Multi-Modal Navigation Systems. Defended June 2026.
MSc, Telecommunications Engineering — 110/110 cum laude
University of Calabria, Italy — ranked 1st in graduating class
Thesis: Wireless Crack Detection Sensor using IoT & Acoustic Emission. Three publications during the programme.
BSc, Electronics Engineering
University of Aleppo, Syria — ranked 1st in graduating class
Telecommunications, electronics, signal processing and embedded systems.
Talks & teaching
IEEE APSCON 2024 — presented “Step Detection Enhanced by Anomaly Filtering for Pedestrian Navigation”
Conference presentations at IEEE I2MTC and IPIN — instrumentation, measurement and indoor positioning
Teaching assistant for antenna design, microwave engineering and RF systems — 50+ sessions, University of Aleppo