Data Scientist; Search & Rankings

About Farfetch

Farfetch is a leading global marketplace for the luxury fashion industry. The Farfetch Marketplace connects customers in over 190 countries and territories with items from more than 50 countries and over 1,400 of the world's best brands, boutiques, and department stores, delivering a truly unique shopping experience and access to the most extensive selection of luxury on a global marketplace.

Our office is near Porto, in the north of Portugal, and is located in a vibrant business hub. It offers a dynamic and welcoming environment where our employees can connect and network with a large community of tech professionals.

We're on a mission to build end-to-end products and technology that powers an incredible e-commerce experience for luxury customers everywhere, understanding the motivations and needs of our customers and partners, to designing and testing hypotheses, to creating industry-leading experiences for luxury customers.

We are seeking a highly motivated Data Scientist to join our Search & Rankings team within the Consumer Products domain. This team is responsible for the core discovery experience that connects millions of luxury fashion lovers with items from over 1,400 of the world's best brands.

You will work in a dynamic, interdisciplinary team alongside Software Engineers and Machine Learning Engineers. While our MLEs focus on building robust MLOps pipelines and scaling infrastructure, your focus will be on the "brain" of the system: deeply understanding user intent, designing complex ranking logic, and proving value through rigorous experimentation.

One of your primary focus areas will be Rankings. Advancing our Learning to Rank (LTR) approaches for Brand, Category, and Search PLPs, which account for approximately 90% of our traffic. Additionally, you will drive the modernization of our Search Engine solutions and broader discovery initiatives as our product scales.

• Algorithm Development:
  Design and develop state-of-the-art Ranking algorithms and NLP models. You will define how products are ordered to maximize relevance and business value.
• Data Science & Strategy:
  Deeply explore our vast datasets (user behavior, catalog metadata) to identify opportunities for personalization. You will answer the "Why" and "What" before we build the "How."
• Experimentation:
  Own the A/B testing framework for ranking/search logic. You will analyze results to distinguish causal impact from noise, ensuring we only ship changes that genuinely improve the customer experience.
• Model Prototyping:
  Build and validate high-quality model prototypes in Python. You will work closely with MLEs to translate these prototypes into scalable, production-ready microservices.
• Collaboration:
  
  Partner with MLEs to ensure your models are compatible with our Databricks/PySpark infrastructure;
  Work with the Catalog Team to define feature requirements;
  Exchange insights with the Recommendations Team (our sister team in Consumer Products) to align on personalization strategies.
• Innovation:
  Stay updated with scientific advancements in Information Retrieval (IR) and Machine Learning, bringing fresh ideas to the table.

• A graduate in Machine Learning, Information Retrieval, Data Science, Computer Vision, NLP, or related fields.
• Algorithm Mastery:
  You have a solid understanding of Learning to Rank (e.g., Lambda
  
  MART, Rank Net) and Information Retrieval techniques. You should have deep expertise in the search domain, spanning traditional methods like BM25 to modern Deep Learning approaches, including Transformers architectures, Sequence Modeling, and Bi-Encoders.
• Python Stack: A strong expert in Python for Data Science (Pandas, Scikit-learn, PyTorch/Tensor Flow, PySpark).
• Data Fluency:
  Able to query and analyze complex data. Familiar with SQL and big data stores (i.e., Big Query, ADLS, and Spark SQL), essential for gathering your own training data.
• Engineering Awareness:
  Comfortable writing clean code that can be easily handed off to MLEs. Experienced in microservices (FastAPI/Flask) is a strong plus. Experienced in Elasticsearch or Solr is also a plus.
• Scientific Mindset:
  You rely on…