Data Scientist

About Topsort

Topsort is a global Commerce AI company making advanced advertising and AI-powered technology accessible to businesses of all sizes.

Our technology helps retailers and marketplaces grow through smarter advertising, real-time auctions, and intelligent product discovery. With a presence in 40+ countries and a growing international team, we're building the technology behind the next generation of commerce.

What Is This Role Like?

We're hiring a Data Scientist to help improve how Topsort ranks and retrieves products for search and ads, predicts user behavior, and supports advertiser planning through reliable predictive models.

You'll work closely with Product and Engineering to turn complex data into actionable models and insights that drive better decisions and continuously improve our platform.

What You'll Own

• Forecasting Models: Develop and refine models for budget planning and advertiser spend, using time-series methods, regression, and machine learning.
• Retrieval: Build and improve the models that select which products are candidates to be shown for each search or placement.
• Ranking: Develop and iterate on rankers and re-rankers that combine relevance, CTR/CVR predictions and business constraints for organic and sponsored placements.
• Query & Product Understanding: Improve how we represent queries and products, including multilingual catalogs, sparse or noisy metadata, and new products with no interaction history (cold start).
• Data Analysis: Explore large datasets to identify trends, assess data quality, and uncover opportunities to improve predictions.
• Production Delivery: Ship models into our ad engine and serving stack, and build the training and evaluation pipelines on large-scale event data.
• Cross-Functional Delivery: Partner with Product and Engineering to integrate models into platform workflows and translate findings into actionable recommendations.

What Success Looks Like

• Reliable Predictions: Models demonstrate consistent performance against defined baselines, offline and in live experiments, with uncertainty and limitations clearly understood.
• Better Relevance: Ranking and retrieval improvements produce measurable lift for shoppers and advertisers.
• Better Planning: Budget and advertiser spend forecasts provide useful inputs for planning and decision-making.
• Practical Adoption: Product and Engineering teams can use model outputs effectively in the platform and relevant workflows.
• Continuous Improvement: Model errors, experiment results and changing data patterns lead to clear, evidence-based improvements.

What You'll Bring

Don't check every box? We'd still love to hear from you. We value curiosity, strong judgment, and people who take ownership of results:

• Relevant Experience: 5+ years of experience in data science, with a strong focus on forecasting.
• Academic Background: A Ph.D. in Data Science, Statistics, Mathematics, or a related quantitative field.
• Ranking & Retrieval: Direct experience building search, recommendation or ads ranking systems in production.
• Deep Learning: Hands-on experience with PyTorch or similar, including training embedding-based retrieval or ranking models.
• Python & SQL: Strong Python and SQL; comfortable building datasets from raw event logs at scale.
• Experimentation: Solid statistics fundamentals and experience designing and reading A/B tests.
• Shipping Mindset: Comfort with AI-assisted development tools and a bias toward getting models into production.

Nice To Have:

• Retail media / AdTech experience, auction or bidding systems.
• LLM-based retrieval, model distillation, approximate nearest neighbor search (HNSW, FAISS).
• Learning-to-rank at scale, or multi-objective ranking.