Senior Engineer Recommendation Systems

About CRITEO

Criteo (CRTO ) delivers personalized performance online marketing through machine learning at an extensive scale. Measuring return on post-click sales, Criteo makes marketing campaign ROI transparent and measurable. Criteo has over 2,200 employees. Engineering HQs are in Palo Alto, California and Paris, France. Our customers include 15,000 advertisers worldwide. Our partners are over 18,000 publishers.

Your Mission:

• Disrupt the state of the art of recommender systems
• Architect, design, and build next-generation algorithms for our recommendation engine.
• Contribute to our strong scientific and engineering culture, with a passion for quality and perfection.

Where are the Challenges?

• Real-time prediction: Everybody predicts clicks. But how do you accurately predict if the user’s click on a product will generate a sale in less than 10 milliseconds? Thankfully, you have billions of data points to help you. But you will always want more, we guarantee.
• Real-time signal: Traditional recommender systems for Big Data are based on batch computation. Can you create a system that updates the recommendations in real-time, based on the user’s latest and current activity on shopping websites?
• Precision: How do you build a high precision recommender system? Increasing coverage is great, but what happens when your ad can be confused with organic results?
• Cold-start: What happens when most of the catalog is made of newly added products?
• Offline testing: You can always compute the offline error on models predicting the probability of a recommendation to convert. But is this really related to the online performance of a new model?

What you bring to the role:

• 10 years of programming experience.
• Good understanding of state-of-the-art supervised and unsupervised learning methods such as logistic regression, factorization machines, and matrix decomposition techniques, and implementing them in a production environment.
• Experience with taking initiative and ownership, exhibiting technical leadership and mentoring skills.
• Experience with developing and extending large-scale and complex recommender systems.
• A rock-solid foundation in Computer Science (data structures, algorithms, software design).
• A passion for shipping quality code.
• Experience with Big Data and technologies such as Hadoop, MapReduce, Spark, Hive.

Preferred/Nice-To-Have Experience:

• Up-to-date on the state-of-the-art topics related to Machine Learning, Recommendation Systems (NIPS, KDD, RecSys...etc): Have you ever published or given a talk?
• Experience starting new products at the ground floor.
• Working in a very fast-paced and continuously evolving environment.
• Have you ever designed your own high-performance, large-scale recommender system?
• Experience in working in a multicultural environment.

Technical Stats about Criteo:

• One of the largest Hadoop clusters in Europe with 40PB of stored data and 3.6PB of data processed every day – comparable to Netflix.
• Analytics Infrastructure comparable in size to Uber or Airbnb.
• Excellent Scalability: 1) 30B HTTP requests and close to 4B unique banners displayed per day; 2) 3M HTTP requests per second handled during peak times; 3) 500B log lines processed per day * 90Gbps of bandwidth, half of it through peering exchanges. We see ~4B cookies/devices per month, corresponding to more than half of the overall Internet population.

Why work at Criteo?

• Career advancement with global mobility opportunities.
• Competitive compensation.
• 100% premium coverage of Healthcare Plan.
• 401(k) fully vested company matching.
• Generous Vacation policy.
• Generous Paternity and Maternity Leave.
• Optional Trips to Paris, France for short-term projects.
• Happy hour, free massages, snacks, arcade/game rooms.

Here is more about Criteo R&D:

A Day in the Life of a Criteo Engineer

Criteo in Palo Alto

Glassdoor

Our Blog:

*Criteo is an equal opportunity employer.

*LI-TD1