Data Scientist, Buyer

About Vinted

Vinted’s purpose is to make second-hand shopping the preferred choice, and the company is seeking people who are excited to help make that vision real. Its teams work to help members buy and sell pre-loved clothing and lifestyle goods so each item can be used again, and again.

The Vinted Group includes three business areas that support this mission: Vinted Marketplace, Vinted Go, and Vinted Pay. Vinted Marketplace is Europe’s leading platform for second-hand fashion and a destination for many categories of pre-loved items, connecting millions of members across more than 20 markets. Vinted Go improves shipping through a network of more than 500,000 pick-up and drop-off points and partnerships with 60+ carriers across Europe, with added services such as item verification. Vinted Pay focuses on secure and dependable payments for buyers and sellers, fully integrated into the app to make transactions safe and efficient.

Founded in 2008 in Lithuania, Vinted started as a simple way for friends to give unwanted clothes a new home. In 2019, the company became Lithuania’s first unicorn. Its headquarters are in Vilnius, and it now operates offices across Europe with a team of more than 2,000 people.

About the Buyer Domain

The Buyer Domain is focused on making search, discovery, and purchase experiences smooth, enjoyable, and highly engaging for members. The challenge is unusually complex because the catalog is peer-to-peer, contains hundreds of millions of unique one-off listings, and receives millions of new uploads every day. Matching the right buyer with the right item in real time is central to the company’s success.

The team is hiring data scientists for two specialist, machine-learning-driven areas:

• Search Relevance: This group builds the retrieval and ranking systems behind search across all markets. Its work covers multi-stage ranking pipelines, query intent classification, semantic dense retrieval, and ML relevance models integrated with content-policy controls for real-time performance.
• Recommender Relevance: This group owns the recommendation system behind the homepage feed and a major share of discovery and purchase activity. Its scope includes deep sequence-based user modeling, real-time candidate generation and ranking, and exploring a broad mix of member interests at very large scale.

The final team assignment will be based on your background, technical strengths, and interests. In either team, you will work closely with product partners, technical peers, and stakeholders, and your work will be measured through offline evaluation and online A/B testing using an in-house experimentation platform.

What you will do

• Design, build, and improve end-to-end machine learning systems for retrieval, ranking, and personalization, including gradient boosting, two-tower models, sequence models, and neural rerankers.
• Work with platform and software engineers to run models inside search and recommendation systems while meeting strict throughput and low-latency requirements.
• Take part in or lead collaborative design discussions for new modeling ideas, features, and training pipelines.
• Set tracking needs, run offline model assessments, and plan and interpret online A/B tests to understand business impact.
• Partner with Data, Engineering, and Product teams to find opportunities, convert business problems into ML goals, and weigh technical trade-offs.
• Contribute to peer reviews, share expertise, and help develop other scientists within the ML Guild and the wider Data Science & Analytics team.

What you should bring

• Hands-on experience designing, training, and deploying machine learning models in production, preferably in search, recommendations, or large-scale personalization.
• Strong Python and SQL skills, plus working knowledge of deep learning frameworks and gradient boosting tools.
• A good understanding of modern information retrieval, learning-to-rank methods, semantic search, or sequence-based preference modeling.
• Exposure to, or strong interest in, high-performance model serving; experience with tools such as Vespa, Elasticsearch, or Triton is an advantage.
• Comfort working with very large datasets in modern cloud-based data warehouse environments.
• Solid knowledge of statistics, offline evaluation metrics, and online experimentation/A-B testing.

Benefits and perks

• Eligibility for the company share options programme.
• 30 days of paid annual leave.
• Latest MacBook models.
• Digital mental and emotional health support, plus an Employee Assistance Program (EAP).
• Home office support, including IT workstation equipment and up to €540 for furniture for a home workspace.
• A lunch benefit for each working day.
• Regular team-building events.
• A personal monthly shopping budget for Vinted.
• Discounted gym membership access.
• A pension plan with company matching at 150% of your chosen contribution.
• Supplemental private health insurance.
• Life and disability insurance.
• A subsidized Deutschlandticket for commuting to the office by public transport.
• Up to 90 workation days per year, including up to 21 days outside the EU.
• A dog-friendly office.

Working at Vinted

Individual Learning Budget: The company supports ongoing growth with an annual budget for courses, certifications, workshops, and similar development opportunities.

Hybrid Work: Vinted uses a hybrid model. Two office days per week are recommended, though not mandatory, and the exact schedule is agreed with your team.

Equal Opportunity: Vinted is committed to an inclusive workplace where people from all backgrounds, identities, and life experiences feel welcome. Applications are considered fairly without regard to race, age, religion or belief, sex, national origin, citizenship, gender identity, sexual orientation, disability, or any other protected characteristic.

Compensation

The annual gross salary for this role is between €61,200 and €100,100.