Freelancer profile translated to English.
Back to original languageAbout Souheil
AI & Data Engineer in Lyon – Specializing in Python, FastAPI, OCR, LLM, NLP, Computer Vision, Cloud Provider (Azure, GCP, AWS...).
AI and Data Engineer passionate about creating end-to-end intelligent solutions, I design, deploy, and maintain complete AI applications, from R&D to production.
I help companies automate their processes, leverage their data, and industrialize their AI models through robust and scalable architectures based on FastAPI, SQLModel, Docker, GCP, Azure, and GPT/LLM.
Recent projects developed at IALab (Lyon):
• Detect’Fraude – Invoice fraud detection (OCR, FastAPI, Azure, LLM, PostgreSQL, SQLAlchemy)
• César & Brutus – Real estate statistics and indicators (FastAPI, SQLModel, Grafana, PostgreSQL)
• GP-Transfer – Document classification and extraction (Factur-X, FEC, Docker, LLM)
• Melana – Administrative automation (Make, Mistral OCR, GPT, Google Sheets)
• Whise Vision – Real estate image analysis (YOLO, Computer Vision, GCP)
I handle all stages: design, development, MLOps, cloud deployment (Azure/GCP), monitoring, and operation, to ensure stable, maintainable, and high-performing solutions in production.
Main Stack: FastAPI · Python · SQLModel · Docker · GCP · Azure · Streamlit · OCR · NLP · LLM · YOLO · Computer Vision · SQL · PostgreSQL · MySQL · SQLAlchemy
📍 Based in Lyon, available remotely.
French
Native or bilingual
English
Fluent
Arabic
Fluent
Remote only
Primarily works remotely
Experience
- IALabData Scientist / AI EngineerDIGITAL AND ITNovember 2024 - Today (1 year and 7 months)Lyon, FranceAt IALab, I work as a freelance AI/ML Engineer to design and deploy custom business AI solutions (NLP, OCR, computer vision, generative AI), enabling the automation of critical processes and delivering real return on investment.I managed the complete AI model engineering: scope definition, data collection and annotation, training and fine-tuning of pre-existing or specialized models (transformers, YOLOv8 detectors, text generators), up to production deployment.To expose these AI capabilities, I structured and designed high-performance APIs with FastAPI, SQLAlchemy, and SQLModel, connected to relational databases as needed (PostgreSQL, MySQL, etc.).Deployment is handled via scalable and containerized cloud architectures: I orchestrated Docker environments, deployed on Azure, GCP, and AWS according to client constraints, while integrating MLOps practices (CI/CD, monitoring, model and data versioning).Beyond pure development, I actively participated in agile AI project management, collaborating with business stakeholders to define key success indicators (KPIs), structure sprint iterations, and ensure scaling.This role allowed me to develop multidisciplinary expertise: advanced Python, Data engineering, Machine learning & deep learning, NLP & OCR, computer vision, cloud computing, MLOps... With a constant goal: transforming the promise of AI into tangible value for businesses.
- EDFAI EngineerENERGY AND UTILITIESApril 2024 - October 2024 (6 months)Nanterre, France• Study and design of an AI chatbot architecture intended to extend SAVE service offerings to the group's various business units.• Exploration and integration of Dialogflow CX (Google Cloud Platform) for creating intelligent chatbots based on decision trees and NLP models.• Development of backend functionalities in Python and Google Cloud Functions, enabling dynamic conversation management and interconnection with internal systems.• Implementation of a secure and scalable cloud environment (GCP, IAM, Cloud Storage) for processing and analyzing conversational data.• Writing technical documents detailing the architecture, permission management, data security, and Cloud & AI best practices.
- Tetra CorpAI Engineer Freelancer - Object RecognitionTECHJanuary 2023 - February 2024 (1 year and 1 month)Lyon, FranceAs part of Epitech's EIP, I participated in the design of Tetra (TRIA), an intelligent software photo library leveraging the latest advances in artificial intelligence and computer vision for object detection and classification in images.My role involved developing and training Python-based Deep Learning models using YOLOv8 (Ultralytics), capable of automatically identifying various types of visual elements (faces, people, landscapes, objects, etc.) with high accuracy.I designed a complete AI pipeline integrating data preprocessing, real-time inference, JSON prediction generation, and model management with DVC (Data Version Control). The development environment relies on MLOps best practices: unit tests (Pytest), static code analysis (Pylint), automatic formatting (Black, isort), and reproducible version management.Several specialized models were trained and optimized (yolo_detect_face_v1_best.pt, yolo_detect_men_women_v1_best.pt, yolo_detect_paysage_v1_best.pt) according to use cases, ensuring fast and reliable detection.This project allowed me to strengthen my skills in Deep Learning, Computer Vision, Machine Learning applied to images, data processing, AI model industrialization, and Python software architecture, while working in a structured and modular environment oriented towards Cloud and production.
Recommendations
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Expert in Information Technology. (Master's degree)EPITECHGrâce à la pédagogie particulière d'EPITECH, les étudiants ont la possibilité d'être au contact direct des entreprises et du monde du travail. En effet, au sein d'EPITECH nous sommes amenés à réaliser des projets en lien direct avec ce qui est d’actualité comme l’intelligence artificielle ou le DevOps, et ce qui est demandé dans le milieu professionnel.