Project Manager in Sber's Anomaly Detection Factory team: managing the development and scaling of an AI product, coordinating cross-functional teams at the intersection of business, finance, DS, and IT.
Tasks:
- Coordinate cross-functional teams at the intersection of business, finance, DS, and IT
- Define key success metrics and target performance indicators
- Connect business results and technical depth: focus on measurable impact, loss prevention, cost savings, and product margin growth
- Formulate and maintain a detailed product backlog, prioritize hypotheses based on business value, technical complexity, and user significance
- Promptly test hypotheses and adjust course
- Generate and test hypotheses based on data, seek insights
- Understand how ML models work with client data
- Research modern approaches to recommendations, including in collaboration with Sber's partner research institutes
- Update the AI product strategy for the Factory
- Deeply immerse into "Lend and Save" products (deposits, loans, auto, commission-based products, BNPL)
Requirements:
- Proficiency in SQL and basic programming languages (Python) for automation of checks
- At least one year of experience as a Product Owner
- Good communication and presentation skills
- Understanding of the full AI model development cycle
- Understanding of principles for working with hypotheses in the field of artificial intelligence
- Ability to work with data: assessing product effectiveness based on metrics, A/B testing
- Experience in evaluating the quality of results (DS and business validation)
Conditions:
- Work format: office/hybrid
- Stable salary and employee social support
- Official employment
- Extended voluntary medical insurance from day one and preferential insurance for loved ones
- Corporate training at Sber's Virtual School
- Prime subscription with the ability to share with three close contacts
- Corporate pension program
- Flexible discount on mortgage loans, equal to ⅓ of the Central Bank's key rate
Skills:
- SQL
- Python
- Spark
- Hadoop
- A/B testing
