ML Product Manager at VKontakte, Moscow
The VK advertising technology team is looking for an ML Product Manager who will be responsible for the creation, development, and scaling of a machine learning (ML) based advertising model. This model will be used in recommendation systems: feeds, clips, Discover, and other surfaces — ensuring ad relevance, high ROI for the business, and user comfort.
Responsibilities
- Formulate the strategy and roadmap for the advertising product in VK recommendation systems
- Manage the lifecycle of ML recommendations — from hypotheses to A/B tests and scaling
- Balance business goals (revenue, CTR, LTV), user experience (engagement, satisfaction), and advertising ethics
- Analyze metrics: impressions, clicks, conversions, ROI, user satisfaction
- Evaluate A/B tests and consider long-term effects, not just short-term results
- Work with attribution models and uplift models to measure the real value of advertising
- Coordinate with teams: analysts, ML engineers, designers, commercial and product specialists
- Formulate technical specifications (PRD) and manage the task backlog
- Prepare reports for stakeholders with transparent analytics and recommendations
- Study the market, competitors, and trends in advertising to find insights
- Formulate data-driven and market-based hypotheses and test them in experiments
- Work with a two-sided ecosystem, considering the interests of advertisers and users
- Think strategically: form a long-term vision for the advertising model in 2–3 years and decompose it into stages
Requirements
- At least 3 years of experience as a Product Manager / Product Owner, with at least 1–2 years working with ML products
- Understanding of machine learning principles, including CTR models, collaborative filtering, and evaluation metrics (AUC-ROC, Precision, Recall)
- Experience in advertising, recommendation systems, or two-sided marketplaces is a strong advantage
- Ability to work with data: understanding of A/B testing, metric analysis, attribution
- Deep understanding of business processes related to advertising: sales, commerce, KPIs, ROI
- Ability to formulate technical specifications and interact with DS and ML engineer teams
- Analytical thinking, ability to find root causes, critically evaluate data
- Communication skills, ability to explain complex things in simple terms and persuade stakeholders
Will be a plus
- Ability to write SQL for data analysis
- Higher education in computer science, mathematics, economics, or related fields
- Portfolio with case studies (previous hypotheses, experiments, results)
- Knowledge of the VK ecosystem, its audience, products, and advertising mechanisms
We offer
- Flexible working hours
- Bonuses and discounts from partners
- Office in the city center
- Voluntary medical insurance (DMS)
- Professional team
Work format
Level
senior
Work schedule
full-time
VKontakte
One of the most high-load projects in the Russian internet. We have up to 30 releases and 125 billion API requests per day. All this is handled by a small team of developers, where each person is a super-professional who has found interesting tasks for themselves.
