Elyes Khechine

• Accomplished 100% success rates in RL-controlled point-to-point contraint-compliant trajectories under industrial-grade timesteps and tolerances through a fine-tuning curriculum of key algorithm parameters • Achieved 96% time-optimality compared to the Ruckig planner and demonstrated model robustness to stricter tolerances and non-standstill speeds via Hindsight Experience Replay, fulfilling the study objectives

• Leveraged a Crystal Agile methodology and a value-driven workflow strategy to lead the team through 5 projects, 6 international competitions and 5 workshops • Streamlined project delivery and ensured adherence to ISO 26262 functional safety standards by designing development V-model lifecycles based on subsystem engineering and HIL testing • Initiated the development of a Qt-based motion-planning simulator for space robots, leveraging Docker to ensure reproducible, end-to-end cross-platform testing

• Developed C++ embedded charging firmware with an ESP-WROVER-KIT, including algorithms for power measurement and dynamic charging speed control • Initiated a production API by prototyping core modules of a Kotlin-based backend server including OTA updates via TCP/IP and load balancing for EV fleet charging • Enabled real-time remote charging control with scheduled data polling and telemetry logging via InfluxDB by integrating a full-duplex WebSocket layer between the cloud server and a user dashboard

• Contributed to the development of 7 out of 42 AI projects, leveraging diverse deep learning and computer vision tools. • Contributed in 70+ knowledge exchange sessions with 15+ AI enthusiasts. Contributed to workshops in machine learning, computer vision, and reinforcement learning. • Led a technical workshop in computer vision and reinforcement learning for robotics, with a focus on policy gradient methods, Q-learning, and model-based approaches.

• Accelerated the design of a 7.4kW AC charger by advising project managers on EV charger architectures for markets in Tunisia and the NL • Collaborated with a team to architect a high-level block diagram of electrical harness between the grid, the charger and the EV, ensuring full compliance with UL and IEC standards • Ensured low-level communication and energy metering with an Infineon MCU, and enabled vehicle presence and crash detection via an adequately chosen MEMS PR accelerometer and AMR magnetometer