Research Scientist - PhD candidate - Next-Generation Generative Models for Search and Recommendation

What if, in your next adventure, you were surrounded by people who, like you, look for an unlimited playground to explore, share, and test, would you care to hear more? You've opened the right door! As an R&D team, making sure your ideas are heard and encouraged is what we strive to do!

What You'll Do:

Criteo AI Lab, operates within the spectrum of two main areas: ML Engineering and ML Research. The position is opened for working in the research department. Research here is mainly dedicated to machine learning topics. The analysis of semantic data has become a major R&D topic for a variety of Criteo business. The PhD offer targets recent advances in the domain of generative deep learning.

PhD proposal : Next-Generation Generative Models for Search and Recommendation

Description

Search and recommendation are at the core of Criteo's business. Generative Information Retrieval (GenIR) and Generative Recommendation (GenREC) are emerging paradigms based on foundation models, promising to transform how we search for and access information. GenIR integrates all components of traditional IR systems into a single generative model, directly generating relevant responses—such as document identities—from user queries. This approach eliminates the conventional distinction between the source of knowledge and the search engine. GenREC operates in a similar manner. Given this framework, search and recommendation—both critical problems for Criteo—can be structured in similar ways. Search becomes the process of matching user queries with catalog products, while recommendation involves matching a user's history and profile with relevant products in the catalog. Looking ahead, GenIR and GenREC will serve as a bridge to LLM-based text generation applications, akin to the role foundation models play today. LLMs are emerging as a new way for users to access information (e.g., on the web or in retail catalogs) and may soon replace traditional search engines and recommendation interfaces. Therefore, developing generative search and recommendation capabilities within LLM-driven interfaces (such as chatbots) is a crucial challenge for Criteo.

The goal of this PhD project is to explore this research direction. The first step will be to develop a unified generative engine for both search and recommendation, allowing for seamless alternation between the two modes during interactive sessions using a single engine. This is also a step toward realizing foundation models that offer a variety of functions to enhance user interactions. The second step will involve adapting this model to the large-scale, dynamic corpora characteristic of recommendation systems in the adtech industry, which presents additional research challenges.

Who You Are:

• You hold a master's degree in computer science or applied mathematics, and/or a diploma from an engineering school.

• You have a strong background in machine learning—both theory and practice—with hands-on experience in deep learning projects and, ideally, large-scale experiments.

• You possess solid technical programming skills.

We acknowledge that many candidates may not meet every single role requirement listed above. If your experience looks a little different from our requirements but you believe that you can still bring value to the role, we’d love to see your application!​

Who We Are:

Criteo is the global commerce media company that enables marketers and media owners to deliver richer consumer experiences and drive better commerce outcomes through its industry leading Commerce Media Platform.

At Criteo, our culture is as unique as it is diverse. From our offices around the world or from home, our incredible team of 3,600 Criteos collaborates to develop an open and inclusive environment. We seek to ensure that all of our workers are treated equally, and we do not tolerate discrimination based on race, gender identity, gender, sexual orientation, color, national origin, religion, age, disability, political opinion, pregnancy, migrant status, ethnicity, marital or family status, or other protected characteristics at all stages of the employment lifecycle including how we attract and recruit, through promotions, pay decisions, benefits, career progression and development. We aim to ensure employment decisions and actions are based solely on business-related considerations and not on protected characteristics. As outlined in our Code of Business Conduct and Ethics, we strictly forbid any kind of discrimination, harassment, mistreatment or bullying towards colleagues, clients, suppliers, stakeholders, shareholders, or any visitors of Criteo. All of this supports us in our mission to power the world’s marketers with trusted and impactful advertising encouraging discovery, innovation and choice in an open internet.

Why Join Us:

At Criteo, we take pride in being a caring culture and are committed to providing our employees with valuable benefits that support their physical, emotional and financial wellbeing, their interests and the important life events. We aim to create a place where people can grow and learn from each other while having a meaningful impact. We want to set you up for success in your job, and an important part of that includes comprehensive perks & benefits. Benefits may vary depending on the country where you work and the nature of your employment with Criteo. When determining compensation, we carefully consider a wide range of job-related factors, including experience, knowledge, skills, education, and location. These factors can cause your compensation to vary.