Senior Data Scientist
HaystackAbout the role
A leading digital media and internet company, focused on content and digital commerce, is seeking a talented individual to shape the future of personalized product discovery. This company connects millions of users with products they love across thousands of retail partners, leveraging a rich understanding of user tastes to create deeply personalized shopping experiences.
The Role
Own the design, build, evaluation, and continuous improvement of recommendation engine models for personalized product feeds.
Take recommendation problems from raw data to production models running on an existing MLOps stack, focusing on the model layer.
Design cold-start strategies to produce high-quality recommendations from small behavioral datasets and growing catalogs.
Define and measure personalization effectiveness through rigorous offline evaluation and online A/B/multivariate experimentation.
Develop deep intuition for product catalogs and user signals, turning implicit behavior and catalog attributes into meaningful model features.
Drive full-cycle ownership, deploying production-quality code, monitoring model performance, and collaborating with MLOps, engineering, and product teams.
What You'll Need
Master's degree or higher in Computer Science, Statistics, Machine Learning, Applied Mathematics, or a related quantitative field, or equivalent practical experience.
Strong data science fundamentals: statistics, experimental design, and evaluation methodology.
Demonstrated ownership of the full A/B testing lifecycle, from design to decision-making.
Experience designing, training, and deploying embedding models and vector retrieval for product or content similarity at scale.
Direct experience with cold-start/sparse-signal personalization for new users or catalogs.
Strong Python proficiency and experience with modern ML frameworks (PyTorch, TensorFlow, or JAX) and the standard scientific stack (pandas, NumPy, scikit-learn).
Strong SQL skills for querying large datasets in a cloud data warehouse (BigQuery preferred) to assemble training/evaluation datasets.
Experience deploying and serving models on a cloud ML platform (Google Cloud Platform Vertex AI strongly preferred).
Strong commerce intuition and understanding of merchandising, category, and product management concerns.
Curiosity and pragmatism regarding emerging AI, particularly LLMs and modern retrieval/ranking, with a track record of bringing new techniques into production.
Excellent written and verbal communication skills to explain technical tradeoffs to both technical and non-technical stakeholders.
What's On Offer
The opportunity to shape how millions of users discover products they love.
A foundational role in building a next-generation product discovery platform.
Work with a company that values rigorous experimentation and data-driven decisions.
Flexible work environment with remote options.
Apply via Haystack today!
Responsibilities
- Own the design, build, evaluation, and continuous improvement of recommendation engine models.
- Take recommendation problems from raw data to production models.
- Design cold-start strategies for high-quality recommendations.
- Define and measure personalization effectiveness through rigorous evaluation.
- Develop deep intuition for product catalogs and user signals.
- Drive full-cycle ownership, deploying production-quality code.
Qualifications
- Master's degree or higher in a quantitative field or equivalent experience.
- Strong data science fundamentals: statistics, experimental design.
- Demonstrated ownership of the full A/B testing lifecycle.
- Experience with embedding models and vector retrieval.
- Strong Python proficiency and experience with modern ML frameworks.
- Strong SQL skills for querying large datasets.
- Experience deploying models on a cloud ML platform.
Benefits
- Opportunity to shape product discovery for millions of users.
- Foundational role in building a next-generation platform.
- Flexible work environment with remote options.
Skills mentioned
About Haystack
Haystack combines AI & expert vetting to deliver world-class tech candidates who are engaged, aligned, and ready to interview. We're trusted by over 400,000+ tech candidates, working in Software Engineering, Data, Design, DevOps, Cloud, Tech Management, Testing, Product & Delivery, Architecture and more. 100s of employers from startups and scale-ups like Atom Bank, DuckDuckGo and Goodlord to established enterprises like American Express, Dunelm and AWS use Haystack to connect with qualified tech talent that they can't find anywhere else.