Bilel Saghrouchni
πLyon | bilel.saghrouchni@gmail.com
Experienceβ
PhD Candidate in Machine Learning β INSA Lyon / SPIE ICSβ
ποΈ Nov 2023 β Present
- CIFRE Thesis: "Anomaly detection through reinforcement and adversarial learning for a Zero-Trust cybersecurity architecture: application in the healthcare sector"
- Research, development, and evaluation of deep reinforcement learning algorithms (Double Deep Q-Learning), machine learning (clustering), and deep learning. Development of a data processing pipeline for model training and visualization. Modeling of a data structure and analysis of network traffic characteristics.
- Collaboration with local hospitals: project management, ML algorithm deployment, data collection, preparation, and visualization for ML applications.
- Communication and popularization of research work and core AI concepts. Participation in R&D strategy around generative AI, testing and presenting SPIE ICS offerings to clients.
- Supervision of an intern (3 months) for the development of a web application for capturing, processing, and visualizing network traffic.
- Stack: Python (TensorFlow, Scikit-learn, Gym, MLFlow, Streamlit), React, Postgres, Docker
Software Engineer β Devoteamβ
ποΈ June 2022 β September 2023
Design and development of innovative solutions around the ServiceNow platform
- Implementation of a Federated Learning algorithm for image recognition in a smart store | TensorFlow, OpenCV, Flask / Python
- Development of a smart voice assistant for technicians | NLP, Rasa, VOSK / Python, JavaScript
- Pre-sales activities: client demos, drafting technical and commercial proposals, participation in calls for tenders
Software Engineer β ETIC, INSA Lyonβ
ποΈ Sept 2021 β April 2023
- Development and deployment of a maintenance application for a national transport company: design and modeling of the database and backend architecture | Git, Angular, TypeScript, Flask, Postgres, AWS BeanStalk, S3, Lightsail
Data Scientist (Intern) β Worldlineβ
ποΈ April β August 2021
- Processing of financial transaction data for fraud prediction: data collection, cleaning, and preprocessing | Pandas, Scikit-learn, Flask
- Optimization of banking transaction data retrieval and development of a web application for integration | Angular, ExpressJS, CubeJS, SQL
Educationβ
INSA Lyon β Telecommunications Engineeringβ
ποΈ 2019 β 2022
- Top 5% of class
- Courses: AI, programming (C++, Python, Go), databases, networks
National University of Singapore (NUS)β
ποΈ January β May 2022
- Academic exchange: Information Retrieval, Computer Security, Feature Engineering
Black belt in Judoβ
17 years of competitive practice: regional and national level. Referee for local-level competitions.
Research and Publicationsβ
- 2025 β Comprehensive Comparison of Streaming Clustering-Based IDS @KES
- 2024 β Towards An Unsupervised Reward Function For A Deep Reinforcement Learning Based IDS @CSNet [DOI]
Technical Stackβ
Projectsβ
- Font Finder: Handwriting font recognition using CNN (TensorFlow, Flask, Streamlit)
- Federated Learning Image Recognition: FL for image classification (TensorFlow, Flask)
- TF-IDF Search System: Optimized text search (Python, nltk)
- Dial-A-Ride Routing Project: Reinforcement Learning (C++)
Awardsβ
- π 3rd place, Code4Good Ericsson (2022), Vivatech Pitch
- π₯ 1st place, Devogame Devoteam (2021)
- π 3rd place, Safran Black Out Challenge (2021)
Last updated: July 2025








