Andrews Cordolino Sobral, Ph.D.
AI Architect & Researcher | 20+ yrs in Software Eng, 10+ in AI/ML/CV | Ex–Head of AI @ ActiveEon | PhD in Computer Vision | Bridging research → real-world AI
Salary / Daily rate
Paris, France
Mobility: Paris
Freelance, open to permanent contract
Can come onsite
Skills
Languages
About me
Intro
AI Architect & Researcher | Former Head of AI @ ActiveEon | Ph.D. in Computer Vision & ML. 20+ years in software engineering and 10+ years in AI & Computer Vision. Expert in AI at Scale, HPC+AI, MLOps, and Computer Vision. 👉 Currently open to senior AI leadership and consulting opportunities. 👉 https://andrewssobral.pages.dev I am an AI Architect & Researcher with 20+ years in software engineering and 10+ years in AI & Computer Vision. As Former Head of AI at ActiveEon, I led a team of Ph.D.-level AI researchers, delivering scalable AI solutions. My expertise spans designing enterprise-grade AI systems, optimizing AI models, and building automation pipelines for large-scale deployments. 🔹 Key Areas of Expertise ✅ AI at Scale | HPC+AI | Distributed & Parallel AI Workflows ✅ MLOps | Enterprise AI Automation | Model Monitoring & Drift Detection ✅ Distributed AI | Multi-GPU & Multi-Node AI Scaling ✅ Computer Vision | Deep Learning | Video Analytics & Anomaly Detection ✅ Embedded & Edge AI | Real-time AI on NVIDIA Jetson, Raspberry Pi ✅ AutoML | Hyper-parameter Optimization | Neural Architecture Search ✅ Generative AI | LLM Fine-Tuning, Deployment & Optimization 🔹 Leadership & Industry Contributions At ActiveEon, I drove AI innovation for industrial and research applications. My main contributions included: • Leading Ph.D.-level AI & ML researchers in enterprise AI development. • Developing ProActive AI Orchestration (PAIO) for large-scale AI workflows. • Architecting AI workflows for multi-GPU, multi-node, and cloud-based AI. • Building AI monitoring tools for real-time observability and drift detection. • Optimizing AI pipelines for fine-tuning, quantization, and efficient inference. • Delivering AI use-cases for clients such as Thales Alenia Space and SAFRAN. 🔹 Research & Technical Contributions I hold a Ph.D. in Computer Vision & ML from University of La Rochelle, France, with research focused on: • Low-Rank & Sparse Matrix Decomposition | Subspace Learning | Tensor Optimization • Real-Time AI for Video Analytics & Edge AI Systems • Open-Source AI Projects – Creator of BGSLibrary (C++) & LRSLibrary (MATLAB) My academic research has led to 2800+ citations (h-index: 17). 🔹 Passion & Side Projects 🚀 Embedded AI & Robotics – Real-time AI for NVIDIA Jetson & Raspberry Pi. 🔍 Open-Source AI – Contributing to AI & Computer Vision libraries. 🔹 Let’s Connect! Open to: ✅ AI architecture & engineering ✅ Computer Vision & embedded AI ✅ Consulting, research collaborations, and advanced AI prototyping ✅ Enterprise AI deployment & automation
Work experience
Associate Researcher (Volunteer, Non-remunerated)
Advanced Institute for Artificial Intelligence
May 2020 - Present
6 years 6 months
São Paulo, State of São Paulo, Brazil
The Advanced Institute for Artificial Intelligence is a consortium of leading researchers from diverse AI fields, dedicated to addressing socially relevant challenges. This is a voluntary, non-remunerated research position within the consortium.
ActiveEon
May 2017 - Jul 2025
8 years 3 months
Paris Area, France
Head of Artificial Intelligence
Jul 2019 - Jul 2025
6 years 1 month
Paris Area, France
Type of contract: CDI (permanent). Leading AI initiatives at ActiveEon, focusing on AI at Scale, HPC+IA, and MLOps. Responsible for developing and orchestrating large-scale distributed AI workflows and automation. Key Responsibilities: • Lead a team of 4 PhD-level AI researchers. • Develop and optimize parallel and distributed AI workflows. • Architect multi-node & multi-GPU AI solutions for enterprises. • Design AutoML frameworks for hyperparameter tuning & neural architecture search. • Oversee AI-driven automation & orchestration for enterprise clients. • Developed & deployed MLOps Dashboard for real-time observability, model monitoring, and performance tracking. • Enabled monitoring & deployment for all AI models, including LLMs & GenAI. • Built a Drift Notification System for proactive model performance monitoring. Key Projects & Use Cases: • ProActive AI Orchestration – End-to-end AI automation platform. • MLOps Dashboard – Centralized AI deployment & monitoring system. • Model Servers & Monitoring – Scalable inference tracking & infrastructure health monitoring. • Drift Detection System – Real-time AI drift & anomaly detection. • LLM & GenAI Integration – Enhanced MLOps for scalable LLMs & Generative AI. • Thales Alenia Space (TAS) – Anomaly detection in satellite hardware. • SAFRAN Aircraft Engines – AutoML-based engine sensor diagnostics. • Hydro Quebec (IREQ) – AI-driven power network analytics. • Desjardins Data Cup 2018 – Led ML model development for financial predictions. Technologies & Tools: • Programming: Python, C++, MATLAB, OpenCV, PyTorch, TensorFlow, Keras, NVIDIA RAPIDS. • MLOps & Deployment: Kubernetes, Apache Spark, Docker, HPC, Distributed AI frameworks. • Monitoring & Visualization: Grafana, Kibana, Prometheus, TensorBoard. • AI Model Serving: Streamlit, Hugging Face, Triton Inference Server.
Artificial Intelligence Research Engineer
May 2017 - Jul 2019
2 years 3 months
Paris Area, France
Senior AI Research Engineer and Machine Learning Team Leader
AI Consultant
ogre.run
Jan 2024 - Dec 2024
1 year
United States
Led the development and optimization of Large Language Models (LLMs) for specialized use cases, focusing on fine-tuning open-source models such as Llama2 7B and Mistral 7B. Key Responsibilities: • Designed AI solutions for real-world enterprise applications, focusing on LLM fine-tuning, model optimization, and efficient AI deployments. • Provided AI consulting to help clients integrate and optimize end-to-end AI workflows. • Developed complete LLM pipelines covering data collection, fine-tuning, evaluation, and model serving. • Implemented parameter-efficient fine-tuning (PEFT) strategies using LoRA. • Optimized inference performance with FlashAttention-2 and AutoAWQ quantization. • Contributed to the open-source project miniogre Technologies & Tools: Hugging Face Transformers · LoRA · SFT · AutoAWQ · FlashAttention-2 · vLLM · Text Generation Inference (TGI) · Docker
Adjunct Lecturer (Vacataire) – Machine Learning and Artificial Intelligence (Practical Approach)
Université de La Rochelle (Master MIX)
Jan 2022 - Dec 2024
3 years
La Rochelle, France
Taught a 10-hour graduate module on Machine Learning and Artificial Intelligence (Practical Approach) [TP] for Master’s students. The course focused on hands-on ML/AI concepts, including supervised and unsupervised algorithms, Python-based implementations, and applied AI project development. Additionally, the module emphasized MLOps, AI in Industry, and AI Automation, exploring how to design, automate, and orchestrate end-to-end AI pipelines — from experimentation to production deployment.
Researcher, Ph.D. Student
Université de La Rochelle
Sep 2013 - May 2017
3 years 9 months
La Rochelle, France
PhD (“European Doctorate” label) on Computer Vision and Machine Learning. Supported by CAPES/Brazil scholarship. Main research tools: Matrix/Tensor Decomposition, Low-Rank and Sparse Representation, and Subspace Clustering.
Computer Vision and Deep Learning Consultant
CamPark Solutions
Jun 2015 - May 2016
1 year
69340 Francheville, France
CamPark Solutions is a French start-up company focused on computer vision solutions for smart city. http://camparksolutions.com/ Type of contract: remote, freelancer I worked as a Computer Vision and Deep Learning consultant in the automated parking lot classification project. My main research activity was to develop a C++ algorithm based on Deep Convolutional Neural Networks (Deep CNN) for image classification. Deep learning tools and libraries: Caffe, NVIDIA DIGITS
Computer Vision Consultant
GeekSys
May 2013 - Oct 2015
2 years 6 months
São Paulo, State of São Paulo, Brazil
GeekSys is a Brazilian start-up company focused on computer vision and big data analytics. http://www.geeksysgroup.com/en/blog Type of contract: remote, freelancer I worked as a Computer Vision consultant in the Heatmaper project: http://www.geeksysgroup.com/en/heatmaper
Software Consultant
XYZTEMAS Consultoria & Servicos
Aug 2010 - Aug 2013
3 years 1 month
Salvador, State of Bahia, Brazil
Software Consultant @ SCAI Project (home-based part-time) SCAI - Sistema Corporativo de Áreas Impactadas Type of contract: remote, freelancer Major Client: Petrobras My activities include: * development of java-based desktop applications with embedded database; * development and testing software components; * mobile apps with HTML5 (websocket, webdatabase, local storage) and Sencha framework; * perform proof of concept; I worked with the following software applications: SCAI2-MEDA - Módulo de Entrada de Dados Ambientais. SCAI2-MEDF - Módulo de Entrada de Dados Financeiros. Tools and technologies employed: HTML5, Java, J2SE, Netbeans, SwingAppFramework + SwingX, JasperReports + iReports, EclipseLink JPA 2.0, H2 Database, SQL Server, Oracle, ...
Software Consultant
GaranteMed
Jan 2009 - Jul 2013
4 years 7 months
Salvador, State of Bahia, Brazil
Jan/2009 - Mar/2010 - Systems Analyst and Software Developer (full-time worker) Apr/2010 - Jun/2013 - Software Consultant (remote, freelancer). The GaranteMed is a Brazilian company of the RedeMed group (medical assurance). Major clients: CPQD, Unimed, Unicred, Uniodonto, Casa de Saúde de Campinas, SENAC and SESC. I worked with software design and software development of web-based applications, and data processing systems. I worked with the following software applications: SGB (Java), GMImport (Java), GMFramework (Java) and SGI (Java), AutorizaWeb (Java), SCGM (Java), GMDesktop (Java), TopSysFramework (Java), GaranteMed (Delphi) and Glote (Delphi). Tools and technologies employed: Java, J2SE, J2EE, Eclipse, Netbeans, Apache Tomcat, Glassfish AppServer, JSP, Servlets, Spring, JSF, Facelets, JDBC, Hibernate JPA, JavaScript, Delphi, SQL Server, HSQLDB, ...
Licenses & certifications
Verified Peer Reviewer
Issued: Dec 2025