Betclic Group has opened applications for a three-year CIFRE Ph.D. position in Data Science based in Bordeaux, France. Offering a permanent CDI contract from day one, the research initiative targets real-time personalization and session-state modeling under strict responsible gambling constraints, bridging academic research with industrial-scale machine learning production.
Bridging Industrial MLOps and Academic Research in Bordeaux
The digital entertainment and sports betting sector moves fast. At Betclic, data isn’t treated as a back-office support function. Instead, dozens of data products and machine learning models run live in production, directly shaping user experiences for millions of players. To bridge the gap between heavy transactional throughput and rigorous academic theory, the company has partnered with the IMS Bordeaux laboratory and ENSC Bordeaux INP for a unique doctoral track.
Candidates selected for the program will enroll through the Doctoral School of Physical and Engineering Sciences (SPI) at the University of Bordeaux. While corporate industrial ph.d. tracks often lean heavily into proprietary commercial optimization, this specific project pivots toward user protection. The overarching research question addresses a core engineering hurdle: how to decode user intent during a live session rather than merely relying on historical preferences, adjusting the platform interface dynamically while prioritizing player safety.
The modern data stack requires sophisticated tooling. Betclic’s data science team operates within a collaborative ecosystem that pairs real-time behavioral telemetry with advanced prototyping toolkits. Notably, the team encourages the responsible deployment of generative AI tools like Claude for research acceleration, code prototyping, and technical drafting.
Deconstructing Session-State Latency and Behavioral Shifts
The core thesis statement assigned to the incoming researcher is titled “De la pertinence de l’item au pilotage de l’expérience : comprendre l’état de session d’un joueur pour personnaliser l’expérience en temps réel, sous contraintes de jeu responsable.” Translating to a shift from static item ranking to holistic experience steering, the research architecture focuses on three explicit technical pillars.
First, the candidate will estimate a latent session state using incomplete event streams. This requires robust handling of behavioral uncertainty and sudden shifts in user engagement regimes. Second, the architecture moves past traditional recommendation filters to design an experience-level adaptation policy. This policy must dynamically weigh multiple interventions: whether to recommend content, provide informational prompts, foster feature discovery, or deliberately suppress further stimulus.
Third, responsible gambling mechanisms are integrated directly into the decision loop as hard constraints. These guardrails rely on early, highly uncertain, and evolving behavioral telemetry observed mid-session. To test these models, the research will evaluate two distinct operational verticals within Betclic’s portfolio: sports betting and online casino environments, utilizing active A/B testing frameworks currently deployed in production.
Requirements and Candidate Profile
Betclic is targeting applicants holding or completing a Master 2 or engineering degree in applied mathematics, statistics, data science, machine learning, computer science, or cognitology. Early-career professionals with one to two years of industrial experience in R&D or data analytics are also encouraged to apply.
The ideal technologist brings a strong foundation in Python and SQL, combined with a demonstrated capability to build experimental protocols and interpret behavioral metrics rigorously. Familiarity with sequential modeling, probabilistic models, latent state estimation, and adaptive systems will prove vital for navigating the dataset scale.
The Industrial and Academic Package
Successful candidates will receive a permanent CDI contract paired with a three-year CIFRE doctoral agreement. Compensation is aligned with junior data scientist salary bands at Betclic, featuring a dual-mentorship structure split between corporate data scientists and academic supervisors.
The recruitment pipeline begins with an initial human resources screening, followed by a technical deep-dive with the Head of Data Science. Shortlisted candidates then complete a scientific defense interview with the academic research supervisors prior to final administrative validation.
