Staff Data Scientist, Trust & Safety
Vinted · Kaunas
Job description
About the role
Vinted's Trust & Safety domain protects its community from harmful activities and content, running machine learning models, rule engines and LLM agents that score listings, accounts and messages. This Staff Data Scientist role shapes the architecture and effectiveness of the detection systems as a whole, combining detection, moderation and sanctioning teams. It is a senior individual contributor position spanning Trust & Safety and Security, with scope to set technical direction, prototype solutions and work with teams to bring them into production.
Key responsibilities
- Design the end-to-end ML architecture for detection, risk scoring, routing and decision-making.
- Develop cost-efficient, layered inference strategies combining broad, low-latency screening with deeper analysis where it adds the most value.
- Establish continuous risk assessment across the member lifecycle, from registration and listing creation to ongoing activity.
- Define how to measure detection effectiveness (precision, coverage, missed harm, computational cost) and use it to drive improvements.
- Turn siloed models into a coherent, calibrated detection system and close coverage gaps.
- Set shared technical standards and architectural direction across teams, working with engineering and data science leaders.
- Prototype critical components, test assumptions empirically and build reference implementations.
- Explain complex ML trade-offs to senior technical and business leaders.
Required profile
- 8+ years of experience in Data Science, Machine Learning or a similar role, including ownership of production models or model evaluation at scale.
- Experience leading cross-team technical work without formal authority and setting standards adopted by others.
- Deep expertise in statistical inference and experimentation: A/B testing, precision and recall under imperfect labels, sampling for ground truth and causal reasoning.
- Understanding of the ML lifecycle in production: evaluation, threshold setting, retraining, drift, feature stores and monitoring.
- Hands-on experience evaluating LLM-based systems, or a strong interest in building rigorous methods for them.
- Fluent English, written and spoken.
- Nice to have: experience in Trust & Safety, fraud prevention, content moderation or risk management.
Required skills
- Highly proficient SQL and Python
- Comfortable in a modern data stack such as BigQuery and dbt
What we offer
- Gross monthly salary of EUR 7,708 to 10,425
- Share options programme and pension plan with a matching contribution
- 25 working days of holiday and a workation policy of up to 90 days per year in the EU
- Hybrid model with two recommended office days a week
- Learning budget starting at EUR 3,000 and a personal monthly budget for shopping on Vinted
- Private health insurance, confidential Employee Assistance Program and dog-friendly office
- Office perks depending on location: gym and in-house meals in Vilnius, lunch allowance in Kaunas
Questions fréquentes
Why are you reporting this job?
Explore further
Salaries, guides and searches in Lithuania.
Apply in 30 seconds
Enter your email to apply. An account will be created automatically.
By continuing, you accept our terms of use.
Already have an account? Login
A question about this job?
Ask it here: you will get the full job summary by e-mail, right away.
Published 1 hour ago
Expires 1 month from now
2 views · 0 interested
Boost your chances
Upload your CV — we will match you with relevant openings.
Analyzing your CV...
Vinted
Kaunas