Master Internship Subject: Developing Tools to Explore and Analyze AI Challenges on Codabench

How about contributing to the development and analysis of a major platform for machine learning
competitions?

Codabench hosts a wide variety of AIArtificial Intelligence challenges, covering different application domains, task
types, datasets, evaluation protocols, and communities. As the platform grows, it becomes
increasingly important to help users discover relevant competitions, understand trends across
challenges, and make better use of the knowledge accumulated on the platform.

As an intern, you will contribute to this project at the intersection of software engineering, data
science, and research on machine learning benchmarks and competitions.

On the engineering side, the internship will focus on improving competition discovery on
Codabench. This may include developing a filtering and search system based on application
domains, task types, data modalities, evaluation metrics, prizes, organizers, activity level, or other
useful criteria. The intern may also contribute to UI/UX improvements, metadata extraction,
tagging tools, and visual dashboards to help users navigate the platform more effectively.

On the research side, the internship will involve a meta-analysis of the challenges hosted on
Codabench. The goal will be to better understand what kinds of AIArtificial Intelligence problems are represented on
the platform: domains of application, types of tasks, evaluation methods, dataset characteristics,
participant activity, and evolution over time. This work may lead to useful insights for the design
of future AIArtificial Intelligence benchmarks and competitions.

The intern may contribute to:

● The design and implementation of a competition filtering system on Codabench
● The definition of a taxonomy for AIArtificial Intelligence challenges
● The analysis of existing competitions and their metadata
● The creation of automatic visualizations and dashboards
● The writing of documentation or reports summarizing the findings
● The preparation of a scientific paper or technical report on the landscape of AIArtificial Intelligence
challenges

Candidates should have a solid background in data science, programming, and Python. Some
experience with Codabench is more than welcome.

The intern may be invited to contribute to the writing of the scientific paper presenting the
platform, or the study produced during the internship.

If you are interested in machine learning competitions and would like to contribute to a major
scientific platform, we would be happy to hear from you.

How to apply?

Please send your CV, a short cover letter, and a link to your GitHub profile to
pavao@mlchallenges.com and acl@lisn.fr

Where?

Laboratoire Interdisciplinaire des Sciences du Numérique (LISN), équipes ASARD et A&O :
1 Rue Raimond Castaing et Rue René Thom (Bât 650 et 660 Digitéo)
91190 – Gif-sur-Yvette

When?

From September to January 2026

Meet the supervisors

• Adrien Pavão, founder of MLChallenges and Codabench main administrator : pavao@mlchallenges.com
• Anne-Catherine Letournel, research engineer at LISN and head of the Codabench project: acl@lisn.fr