About
AI Architect and Researcher | Former Head of AI at ActiveEon | Ph.D. in Computer Vision & Machine Learning
20+ years in Software Engineering • 10+ years in AI, Computer Vision, Deep Learning, and Distributed Systems
I am an AI Architect and Researcher focused on bridging the gap between research innovation and industrial-scale deployment. My career combines deep technical expertise with strategic leadership, specializing in high-performance computing, computer vision, generative AI, and large-scale AI automation.
I design and deploy production AI systems that translate cutting-edge research into real-world industrial applications.
My work sits at the intersection of software engineering, artificial intelligence, and large-scale distributed systems, with a strong focus on AI at scale across cloud, HPC, hybrid, and edge environments.
As Former Head of AI at ActiveEon (2017–2025), I led PhD-level AI researchers and helped transform advanced AI concepts into production-ready platforms for enterprise and research use cases.
With a software engineering foundation dating back to 2001 and specialized AI research since 2010, I bring end-to-end expertise spanning architecture, distributed training, optimization, deployment, and production inference.
Core Expertise
- AI Architecture & Scale — HPC for AI | Distributed & Parallel Workflows | Multi-GPU/node scaling | Federated & hybrid-cloud systems
- MLOps & Automation — Experiment tracking | Continuous Training (CT) pipelines | CI/CD for ML | AI workflow orchestration (cloud, on-prem, HPC) | Monitoring, drift detection, and model governance
- Computer Vision & Image/Video Processing — Object detection, anomaly detection, motion analysis | Real-time video analytics & CV pipelines | Background subtraction, low-rank/sparse decomposition
- Embedded & Edge AI — Real-Time Inference (Jetson, RPi) | TensorRT, OpenVINO optimization | DeepStream, GStreamer | Constrained hardware inference
- AutoML — Hyperparameter Optimization (HPO) | Neural Architecture Search (NAS)
- LLMs & GenAI — Fine-tuning, PEFT/LoRA, Quantization | FlashAttention, AutoAWQ | Deployment (Triton, vLLM)
- Stack — C++, Python, CUDA, AI SDKs, APIs, JupyterLab Integration
Selected AI Use Cases
Smart Transportation & Edge AI Systems
- Designed and deployed real-time vehicle detection pipelines using NVIDIA Jetson Nano, DeepStream, and YOLO for multi-stream RTSP analytics and vehicle counting.
- Built smart parking occupancy systems using IP cameras and edge devices, with end-to-end deep learning pipelines and custom ROI annotation tooling.
- Systems operated continuously (24/7) in real-world production environments with high reliability requirements.
Leadership @ ActiveEon — Former Head of AI (2017–2025)
At ActiveEon, I led PhD-level AI researchers and architected large-scale AI platforms for enterprise and industrial applications, scaling systems from research prototypes to production deployments. My responsibilities included:
- Led PhD-level AI researchers in AI/ML, GenAI, and Vision Systems
- Architected ProActive AI Orchestration (PAIO) for distributed AI workflows across hybrid cloud and on-prem environments
- Deployed LLMs, deep learning, and AutoML pipelines across multi-node, multi-GPU, HPC, and hybrid cloud environments
- Built custom SDKs and tools for dynamic workflow automation (e.g., ProActive Python SDK, ProActive Jupyter Kernel)
- Contributed to the ExtremeXP European research project by leading the implementation of a runtime for scheduling and executing complex analytics workflows on distributed infrastructure using ProActive AI Orchestration (PAIO). This includes integration for distributed AutoML, resource monitoring, and dynamic service orchestration via the ProActive Python SDK, enabling on-the-fly deployment of tools like TensorBoard and MLOps dashboards.
- Designed and deployed AI orchestration workflows for LuxProvide’s MeluXina Tier-1 supercomputer, enabling distributed multi-node/multi-GPU training with PyTorch DDP, large-scale AutoML, MLflow-integrated experiment tracking with SLURM, and high-performance LLM inference (Llama 3.1 405B FP8) using vLLM on Ray clusters
- Developed distributed AI workflows for IDRIS / Jean Zay in partnership with HPE, enabling multi-node/multi-GPU orchestration with MPI, gRPC, Horovod/NCCL, containerized and bare-metal execution, and scalable AutoML-driven hyperparameter optimization with integrated experiment tracking and SLURM-native scheduling
Selected Enterprise & Industrial AI Use Cases
- Thales Alenia Space (TAS) — Developed AI workflows for satellite manufacturing and operations, including defect detection, anomaly detection, and telemetry analysis, reducing per-test analysis time to minutes.
- SAFRAN Aircraft Engines — Built AutoML-driven predictive maintenance pipelines over large-scale sensor data (2,000+ sensors), enabling real-time anomaly detection and rapid engineering insights.
- IRT Saint-Exupéry / TAS — Contributed to autonomous image processing pipelines for aerial and satellite imagery, including semantic segmentation for maritime detection use cases.
- Hydro Québec (IREQ) — Designed ML workflows for power grid optimization, integrating real-time analytics, anomaly detection, and forecasting over streaming infrastructure.
- Desjardins — Developed distributed ML pipelines for predictive analytics on financial data, including client default forecasting over a 12-month horizon.
Academic Background & Research
Ph.D. in Computer Vision & Machine Learning – Université de La Rochelle, France
Specialized in:
- Low-rank & sparse matrix/tensor decomposition
- Subspace learning & optimization
- Multimodal and real-time video analytics
- Embedded computer vision (Jetson, Raspberry Pi, PandaBoard)
My academic research has led to 2800+ citations (h-index: 17) and contributions to peer-reviewed journals and top conferences (IEEE, Elsevier, Springer, CVIU, TIP, TNNLS, ICCV). I also actively contribute as a peer reviewer and open-source developer.
Open-Source & Contributions
🔗 GitHub: github.com/andrewssobral
- BGSLibrary (C++) – Widely used background-subtraction library for moving-object detection.
- LRSLibrary (MATLAB) – Framework for low-rank/sparse decomposition.
- MTT (MATLAB) – Tools for tensor manipulation and decomposition.
- OSTD (MATLAB) – Online stochastic tensor decomposition for multispectral video.
- VDTC (C++ & Python) – Vehicle detection, tracking & counting pipeline.
- GoDec (Python) – Low-rank + sparse decomposition.
- DTT (C++) – Header-only library for seamless data type conversions (Eigen, OpenCV, Armadillo, LibTorch, ArrayFire).
These projects are used in academia, industry, robotics, surveillance systems, and research labs worldwide.