Software engineering and technical leadership in Java/Spring Boot and Angular/TypeScript, with a specialization in backend architecture and performance optimization.
Software Engineer since Nov. 2021 Tech Lead since Oct. 2024
Technical lead within a five-person development team working on an IoT platform used by B2B clients to track assets and optimize their operations.
Guided architecture decisions across the platform and client applications, evolving existing systems incrementally, including the ongoing move to a modular monolith. Documented the architecture and APIs, and wrote specifications for upcoming features.
Scaled backend processing to handle up to 400,000 events per day in production on 2 vCPUs and 8 GB of RAM, and load-tested it for a one-million-per-day requirement. Replaced N+1 queries and repeated database access with batching and caching, cutting processing time from 150–300 ms to roughly 12 ms per event.
Led technical discussions with the team, weighing objections against constraints and trade-offs: adapting proposals where concerns revealed weaknesses, and making the call when views still differed. Helped developers debug difficult issues and apply best practices.
Alongside the main platform, developed separate applications for clients in healthcare and logistics. Worked directly with them to clarify their needs and business rules, agree on the required changes, and ensure compatibility with their existing business systems.
Set up application observability (metrics, error rates and logs with OpenTelemetry, Prometheus, Loki and Grafana), and introduced SonarQube quality checks, stricter ESLint rules including Angular-specific ones, and Docker-based development environments.
Developed Java and Spring Boot modules to integrate IoT devices with varied message formats and uneven protocol standardization. Implemented message encoding, developed SDKs and handled remote firmware updates.
Built Angular interfaces for exploring device data and history through interactive maps, charts and customizable tables. Enabled users to inspect records in detail and navigate related information, using OpenLayers for maps and Chart.js for charts.
Evaluated coding agents and established team guidelines for their use in development: which tasks to delegate, what data agents can access, and how generated code should be reviewed.
Built a web application for a statistics team to browse and edit the descriptions of its datasets, with forms generated from their field definitions.
Evaluated Google Earth Engine against the research team's existing tools by reimplementing their processing tasks and benchmarking them.
University of Southern Brittany · Vannes, France
Web and desktop development, distributed computing, machine learning, functional programming and security.
University of Southern Brittany · Vannes, France
University of Southern Brittany · Vannes, France
Built a shared dashboard for following streamers participating in a charity event, showing their status in real time. Developed with React and Node.js.
Bordeaux, France