Data Scientist Recommender Systems

EyeSpy Recruitment; España; Remote; EUR 55,000 - 75,000; Full time

Job summary

EyeSpy Recruitment is seeking a Data Scientist specialising in Recommender Systems for a remote role based in Spain. You will build models that generate personalized recommendations for millions of players and bring them from experiments to production.

Work with Python, PyTorch and deep learning in a modern cloud stack, deploying models, evaluating them offline, and exploring transformer-based or two-tower approaches to boost engagement across games.

Qualifications

• 3–5 years hands-on ML experience with recommender systems.
• Strong Python and SQL skills.
• Experience taking models to production and deployment.
• Familiarity with cloud platforms (AWS/GCP/Azure) is a plus.

Responsibilities

• Build recommendation models for candidate generation and ranking.
• Use sequential behaviour and embeddings for next-item prediction.
• Ship models from experiments to production and monitor them.
• Run offline evaluation and A/B tests to link quality to product results.
• Experiment with transformers, two-tower models and multimodal signals.

Skills

Machine learning

Recommender systems

Personalisation

Python

SQL

Tools

PyTorch

TensorFlow

JAX

AWS

GCP

Azure

Job description

Data Scientist, Recommender Systems | Remote | Python, PyTorch, Deep Learning

Millions of players. Thousands of games. One question: what should they play next?

We're hiring a Data Scientist specialising in Recommender Systems for a fast-growing global gaming technology company. You'll build the machine-learning models that decide what millions of players discover, and you'll see your work go live, not sit in a notebook.

What you'll actually do

• Build recommendation models. Candidate generation, retrieval and ranking that make every player's lobby feel personal.
• Predict what's next. Use sequential behaviour, user and item embeddings and deep learning for next-item prediction.
• Ship to production. Take models from data and experiments through deployment and monitoring, then keep improving them.
• Prove it works. Run offline evaluation and A/B testing, and connect model quality to real product results.
• Try new things. Test transformer-based recommendation, two-tower models and multimodal signals where they genuinely help.

What you'll bring

• Around 3–5 years of hands-on experience in machine learning, with real work on recommender systems, personalisation or ranking
• Strong Python and SQL
• Deep-learning experience with PyTorch, TensorFlow or JAX
• Models you've taken to production on AWS, GCP or Azure
• The ability to explain what you built, why, and what changed because of it

Two-tower models, learning to rank, ANN search, cold start, session-based recommendations, MLOps, Spark, Kafka or Flink. Experience from e-commerce, streaming, media, marketplaces or ad tech counts. Your job title matters less than what you've built.

Why you'll like it

• Real scale: your models reach millions of players worldwide.
• Real ownership: you own problems from idea to production.
• Real growth: learn alongside data scientists, ML engineers and data engineers who ship.
• Modern stack: deep learning, cloud ML and real-time data, with room to test new ideas.