AI Agents Intern: Build Production-Ready Product Tools

Criteo; Paris; Hybrid; EUR 10,000 - 17,000; Full time

Benefits offered by this job

Hybrid work model

Learning & career development programs

Health benefits & wellbeing perks

Diverse, globally connected team

Competitive pay & equity potential

Job summary

Criteo in Paris is offering an internship with the Product Analytics & Data Science team (PAX) starting January 2027. You will design, prototype, and evaluate AI agents and reusable skills that tackle real product and analytical problems, embedding them alongside Product, R&D, GTM and Design teams.

You will build agents, evaluate guardrails, and convert repetitive tasks into scalable workflows, with hands-on exposure to LLMs, embeddings, retrieval, and agent evaluation, under mentorship in a

Qualifications

• Master's degree student or higher in a quantitative field.
• Available for 5–6 months from January 2027.
• End of study or gap year internship.
• Strong analytical and builder mindset with ability to prototype.

Responsibilities

• Design, build and evaluate AI agents and reusable skills for product/analytic problems.
• Frame product/client problems with PM/R&D/GTM/Design and prototype high-value solutions.
• Build evaluators, guardrails and golden-dataset tests for trustworthy systems.
• Turn repetitive tasks into scalable, trusted agents.
• Mine large datasets and translate into actionable insights.
• Master our internal analytics datasets and AI stack.

Skills

Python

SQL

Data analysis

AI agents

Dashboards

Education

Master's degree student or higher

Tools

LLMs

Embeddings

Retrieval

Job description

Criteo in Paris is offering an internship with the Product Analytics & Data Science team (PAX) starting January 2027. You will design, prototype, and evaluate AI agents and reusable skills that tackle real product and analytical problems, embedding them alongside Product, R&D, GTM and Design teams.

You will build agents, evaluate guardrails, and convert repetitive tasks into scalable workflows, with hands-on exposure to LLMs, embeddings, retrieval, and agent evaluation, under mentorship in a