Freelancer profile translated to English.

Description

Data scientist/Fullstack developer. I have studied and have experience in AI. My experience extends to full stack development, but also to AI issues, particularly in NLP, but also in clustering, optimization, recommendation systems, time series, and computer vision. With a strong appetite for Development, I am able to quickly set up POCs as well as complete products, and eventually move towards production and deploy solutions through APIs and services, using technologies for containerization, orchestration, and continuous integration to ensure a robust system and build a solid, scalable, and maintainable architecture.

Curious and passionate, I am always motivated to learn and get my hands on new technical stacks, which has allowed me during my experiences to step out of my comfort zone and develop new skills.

Languages

French
Native or bilingual
English
Fluent
Arabic
Fluent

Workplace preferences

Remote only

Primarily works remotely

XPDEEP
Developer / Data Scientist
SOFTWARE PUBLISHING
August 2023 - Today (2 years and 10 months)
Grenoble, France
• Backend development (from scratch) with Flask - SQLAlchemy - PostgreSQL.
• Dependency management and versioning with Poetry.
• Containerization of backend-related services: backend, database, nginx. Writing Dockerfiles and Compose to orchestrate the different services.
• CI/CD with GitlabCI. Implementation of the necessary jobs: Lint (ruff), unit tests and integration tests (pytest, unitest), static analysis (mypy), build and push of docker images with automated versioning via tag creation, etc...
• Migration from Flask to FastAPI + Pydantic.
• OIDC authentication system and implementation of an FGA for rights management.
• Generation and maintenance of the OpenAPI file, Swagger, and a client for the backend.
• Development of a client (python library) for more ergonomic use of XPDEEP APIs. Implementation of Gitlab CI/CD pipelines for package generation and versioning. Automated generation of client documentation with Mkdocs and its publication.
• Implementation of neural networks (Torch) and training on different types of state-of-the-art datasets (classification, regression, time series forecasting, etc...). Testing APIs via the XPDEEP client by reimplementing the neural network architectures already designed with Torch, in order to evaluate our solution and validate its various functionalities.
• Maintenance and bug fixing in the backend, client, and CI/CD pipelines.
• Implementation of a Loki database and Grafana dashboards for backend log observability.
• Implementation of an Admin Interface for the database.
• Implementation of QA and end-to-end tests.
Docker Gitlab CI Pytorch FastAPI SQLAlchemy
Akkodis Research
Data scientist in computer vision
SOFTWARE PUBLISHING
January 2022 - June 2023 (1 year and 5 months)
78280 Guyancourt, France
• Identification of existing Deep Learning approaches in the literature for heart rate (HR) estimation from video.
• Implementation and adaptation of approaches from research papers for calculating HR (heart rate) via a standard camera (ippg): DeepPhys & BVPNet
• Face detection and region of interest detection from video data
• Visualization, testing, and evaluation of results
Deep Learning Scikit-learn Computer Vision Python Machine learning
Akkodis Research
Full stack developer / NLP Engineer
SOFTWARE PUBLISHING
January 2022 - Today (4 years and 5 months)
78280 Guyancourt, France
• Study of different XAI methods and their implementation
• Proposal and implementation of a new XAI approach, illustrated via a scatter plot
• Deployment of different solutions as APIs
• Unit + integration tests
• Participation in the integration of data-related functionalities on the front-end: front-end integration, bug fixing, etc...
• Migration of the backend from Node (Javascript) to Flask (Python) / Adaptation of the front-end / Implementation of a MongoDB-type database
• User and authentication management
• Project conducted in an agile environment (daily stand-ups, sprint retrospectives, sprint planning)
flask MongoDB Python