Heythem SATTOUTAH
Applied AI & Distributed Systems | Networks, DevOps & Embedded AI | Sorbonne & Renault Group
Salary / Daily rate
91120 Palaiseau, France
Freelance, open to permanent contract
Skills
Languages
About me
Intro
👋 Hi, I’m Heythem (Haithem) SATTOUTAH — graduate of Sorbonne Université (M2 Computer Science) and ESI SBA (Engineer’s degree in Computer Science). I recently completed my R&D internship at Renault Group, where I worked at the crossroads of AI, distributed systems, and networking. I’m passionate about turning advanced algorithms into robust, measurable systems on real hardware. 🚗 R&D @ Renault Group Designed & optimized multi-agent connectivity solutions with clustering to improve reliability & QoS. Migrated LLM inference from Tinygrad → PyTorch, reducing latency on Jetson Nano. Ran distributed training experiments with EXO Gym, benchmarking FedAvg, DiLoCo, DeMo & SPARTA. Built automated pipelines for embedded AI with real-time inference + fallback strategies. 🧠 Research & Publications Modeled p-center facility location problems via Max-SAT at UPJV (MIS Lab). Published at JFPC 2024 & CP 2024, bridging theory & scalable solving. 📦 Key Projects ROADEF 2024 Scheduling (CPLEX) 🧮 – Optimized conference session assignments. Max-SAT Scheduling (PySAT) 🔧 – Encoded constraints, minimized conflicts in large instances. Raspberry Pi + Kafka 📹 – Real-time object detection pipeline with Dockerized deployment. Distributed Training (PyTorch) 💻 – Benchmarked strategies on decentralized setups. BGP Simulation (GNS3 + Wireshark) 🛰️ – Automated routing topologies & monitoring. 🧰 Toolbox 💻 Python, PyTorch, TensorFlow, scikit-learn 🐳 Docker, Linux, Git, CI/CD ⚡ Kafka, MongoDB, MySQL, Redis 📡 GNS3, Wireshark, routing protocols 🧮 PySAT, CPLEX, Max-SAT solvers 📱 Jetson Nano, Raspberry Pi 🌍 Community & Languages Volunteer: AlphaBit Club (Finance) & Algerian Muslim Scouts (Leader & Media Officer) Languages: Arabic 🇩🇿 | French 🇫🇷 | English 🇬🇧 🎯 Next Step Now an alumnus 🎓, I’m seeking opportunities where I can blend AI, Networks, DevOps & Systems to deliver scalable, reliable solutions. Excited by roles in R&D, Embedded AI, Robotics, Distributed Systems, and Data/ML Engineering. If you’re building systems that must run fast, reliably, and at scale — let’s connect! 🚀
Work experience
R&D Engineer : Distributed Hybrid Intelligence: Communication and Clustering Strategies
Renault Group
Feb 2025 - Aug 2025
7 months
78280 Guyancourt, France
During my internship at Renault Group – Ampere (Advanced Research & Technologies), I worked on the design and optimization of hybrid AI and multi-agent connectivity solutions for connected vehicles. Leveraging Exo Labs’ distributed computing framework and the DiLoCo algorithm, I enabled parallel and distributed LLM inference across heterogeneous devices (Jetson edge modules, PCs, smartphones) to maintain low-latency communication and high QoS under variable network conditions. My work combined simulation (Exo Gym), network emulation for QoS testing, and real-vehicle deployment. I benchmarked distributed training strategies (FedAvg, DiLoCo, SPARTA, DeMo), which rely on parallel gradient aggregation and synchronization, recreated real vehicular scenarios (urban handovers, tunnel blackouts, rural coverage), and deployed optimized inference engines (PyTorch vs MLX-CUDA) that exploit GPU-level parallelism in a heterogeneous in-car cluster. This project strengthened my expertise in distributed AI, reinforcement learning, networking, and embedded systems, while contributing to Renault’s vision of building the software-defined, AI-driven vehicle of the future. 🚗⚡🤖
Minimizing Working-Group Conflicts in Conference Session Scheduling Through Maximum Satisfiability
Université de Picardie Jules Verne
Feb 2024 - Jul 2024
6 months
Amiens, France
- Internship focused on the Maximum Satisfiability (Max‑SAT) problem to solve p‑center location problems. Includes literature review, problem modeling in Max‑SAT, and solving them using specialized algorithms. The internship aims for applications in various fields such as planning and facility problems. -Accepted paper at the national conference in France JFPC(Journées Francophones de Programmation par Contraintes). - Accepted paper at international conference CP(Constraint Programming).
Seasonal Team Member
Groupe Amor Benamor
Jul 2023 - Jul 2023
1 month
Skikda, Algeria
Practical internship at Data Center Sonatrach
Sonatrach
Jan 2023 - Jan 2023
1 month
Skikda, Algeria
Practical internship at Sonatrach Ra1k
Sonatrach
Sep 2022 - Oct 2022
2 months
Skikda, Algeria
Seasonal Team Member
Groupe Amor Benamor
Jul 2022 - Aug 2022
2 months
Skikda, Algeria
Education
Sorbonne University
Master's degree, Computer Systems Networking and Telecommunications
2024 - 2025
2 years
École Supérieure en Informatique de Sidi Bel Abbès – ESI SBA
Engineer's degree, Computer Science
2019 - 2024
6 years
Licenses & certifications
Model Context Protocol (MCP): Hands-On with Agentic AI
Issued: Oct 2025
Advanced Learning Algorithms
Issued: Oct 2023