About Sonia
Hi there 👋
I’m an ML engineer passionate about building AI systems that are not just smart, but also clean, scalable, and production-ready. From early-stage prototypes to fully deployed pipelines, I love turning complex ideas into real, impactful products.
With a solid background in both R&D and client-facing projects, I bring a mix of deep technical skills and a strong product mindset. I enjoy collaborating with tech and product teams to co-create solutions that actually work — fast, maintainable, and easy to integrate.
Whether it’s deploying vision models, setting up monitoring for production ML, or making LLMs more useful in real workflows, I’m all about creating value through AI — with clarity and care.
French
Native or bilingual
English
Fluent
Remote only
Primarily works remotely
Experience
- PICSELLIADATA SCIENTIST / MLOPS ENGINEEROctober 2021 - Today (4 years and 8 months)As a full-stack ML Engineer, I’ve contributed to both internal R&D and client-facing projects with a strong focus on model deployment, optimization, and automation.
- Model Integration & Optimization: Integrated and productionized computer vision models in a scalable, Dockerized pipeline with unit/integration testing.
- Plug-and-Play AI Tools: Designed AI-powered tools with clean APIs, ready for seamless integration by back-end and front-end teams.
- R&D Innovation: Built monitoring tools (outlier detection, data drift) and integrated visual-language models to enhance image search. Automated documentation workflows using LLMs.
- Client Project Delivery: Led end-to-end client projects, from needs assessment to delivery. Ran MLOps workshops and managed annotation teams for data labeling.
- Platform Contribution: Performed QA testing for the platform and SDK, and contributed feature requests to improve UX and dev efficiency.
- Documentation: Maintained clear, developer-friendly documentation for both internal tools and client-facing deliverables.
- ID-PalDATA SCIENTIST INTERNJune 2023 - September 2023 (3 months)Worked on a fraud detection use case, applying computer vision to identify forged identity documents.
- Fraud Detection Model: Designed and trained a deep learning model using open-source datasets to detect fake IDs via document analysis.
- Plug-and-Play Deployment: Delivered a fully integrated solution designed to be easily consumed by backend and frontend teams for production use.
- Performance Optimization: Fine-tuned the model for real-time performance and reliability.
- Documentation: Provided detailed documentation to support long-term maintenance and development.
- N7 CONSULTINGTECHNICAL LEADERNovember 2020 - December 2021 (1 year and 1 month)Led technical aspects of client projects within a student-run consulting firm, combining project management with hands-on development.
- Client Interaction: Participated in client meetings to understand needs, propose solutions, and define project scope and budgets.
- Team Management: Coordinated a team of junior consultants, reviewed code, and guided project delivery from start to finish.
Recommendations
Be the first to recommend Sonia
Help this freelancer shine by sharing your experience working together.
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
- Ingénierie : Science du numérique (Informatique/Mathématiques appliquées/ Télécommunications/Réseaux)ENSEEIHT2021Ingénierie : Science du numérique (Informatique/Mathématiques appliquées/ Télécommunications/Réseaux)
- Classes préparatoires aux grandes écoles (CPGE), Mathématiques et physique en classe étoiléeLycée Masséna2020Classes préparatoires aux grandes écoles (CPGE), Mathématiques et physique en classe étoilée