Data Scientist Intern, Back Market

• Mandatory: Currently enrolled in your last year of an Engineering School with specialization in Data Science/Engineering or University Master's degree and your university can provide an internship agreement
• Strong foundation in AI/ML concepts; experience working with Large Language Models is appreciated
• Proficiency in Python and experience with ML libraries
• Analytical mindset with experience in data analysis and statistical evaluation methods
• Passionate about AI innovation and eager to work with cutting-edge technologies
• Strong problem-solving skills with attention to detail in model evaluation and testing
• Our working language is English; fluency is required (no French required)
• Contract: Internship agreement from your French school required
• We know that the perfect background for a role doesn’t mean the perfect fit — we encourage you to apply for a role even if you think you may not have all the qualifications

Desirable

• Join Back Market's AI in Care Team as we accelerate AI adoption across the company and turn conversations into action
• Our team builds Back Market's ML and AI applications in the Customer Care domain
• We've presented our LLM research at several applied AI conferences, showcasing for instance how we blend large language models with traditional ML techniques
• This innovative approach gives us deeper insights into customer interactions, which translates into a meaningfully improved platform experience for our users
• As a Data Science AI Intern, you will focus on understanding and evaluating Large Language Models (LLMs) in one of two key settings, depending on business priorities and your interests
• You will build and run evaluation frameworks, compare models, uncover blind spots and failure modes, and partner with engineers and product teams to translate findings into improvements
• The first setting involves LLMs used as feature extractors in internal tools, where we combine LLM-derived signals with machine learning models to produce actionable KPIs and predictions
• The second setting focuses on LLM-based chatbots, an area we're actively exploring, where you'll develop rigorous evaluation and red teaming approaches to ensure reliability, safety, and the best consumer experience
• This internship is central to our AI strategy, highly visible, and offers a rare opportunity to have transformative impact
• You will contribute to production systems used across the business while learning fast in a hands-on environment
• Develop and implement comprehensive LLM evaluation frameworks to assess performance and reliability across two key applications:
• LLM evaluation in our tech stack: Evaluate the accuracy and consistency of LLMs integrated within our technology and analysis stack, for instance to extract meaningful insights from customer conversations (sentiment, tone, behaviors) or other technical applications
• You'll uncover potential blind spots, analyze failures, and compare different models and parameters to identify which perform best for our specific use cases and how to generalize this approach
• Chatbot Applications: Design thorough testing methodologies to explore edge cases, conduct red teaming exercises, and ensure reliable performance across diverse user interactions and scenarios
• You'll also evaluate implementation considerations including optimal model choice and model parameters, prompt engineering strategies to ensure both safety and customer experience
• Collaborate closely with engineering and product teams to translate evaluation insights into actionable improvements, helping guide model selection decisions and system enhancements
• Support the team's broader ML initiatives by contributing to statistical analysis, data science projects, and other machine learning efforts that align with your skills, interests, and the team's evolving priorities
• Duration: 6 months (until September 2026)
• Schedule: Full-time (35H/week)

• Video-call Interview with the Campus Tech Recruitment Manager (30min)
• Video-call Technical fit interview : Live coding + open talk (60min) with the tutor
• Video-call Team fit interview with 2 engineers/members from the team (30min)
• Starting date: March 2026