This is a chance to start a career at the forefront of artificial intelligence security. You will join a team that researches vulnerabilities and protects GenAI models (LLM/MLLM) and applications. Under the guidance of experienced experts, you will participate in real research and engineering tasks to understand how AI systems are hacked and how they are defended.
Who we are looking for:
Profile and education: senior student or graduate in computer science, applied mathematics, information security, or related fields.
Employment: at least 20 hours per week, flexible schedule (at least 2 days a week in the office)
What you will learn:
- Research: assisting in the analysis of new publications, incidents, and attack methods on GenAI (studying arXiv, blogs, repositories)
- PoC development support: participating in the creation and testing of Proof of Concept (PoC) prototypes to demonstrate vulnerabilities under the guidance of a mentor
- Working with tools: mastering and assisting in improving internal tools for testing the resilience of models and applications
- Experiments: participating in model security assessments — conducting basic tests, analyzing results, preparing data
- Immersion in the ecosystem: actively studying frameworks and libraries for working with AI (LangChain, Transformers, etc.).
What is important to us:
- basic Python programming skills and a desire to write clean, understandable code
- interest and initial understanding of how Large Language Models (LLM) and Multimodal Models (MLLM) work, as well as their associated security risks
- English language proficiency sufficient for reading technical documentation and articles (e.g., on arXiv)
- basic knowledge of mathematical statistics for meaningful participation in experiments.
Will be a significant advantage:
- personal or academic projects related to AI/ML (even simple ones): model fine-tuning, chatbot development, Kaggle participation
- experience (even minimal) working with any of the libraries: PyTorch, transformers, LangChain, scikit-learn
- familiarity with cybersecurity fundamentals and key concepts (e.g., from OWASP Top 10)
- CTF (Capture The Flag) participation, hackathons, or ML/AI courses
- presentation skills and the ability to clearly explain your ideas.
We offer:
- comfortable modern office
- office-based work format
- annual salary review, annual bonus
- corporate gym and recreation areas
- over 400 educational programs from SberUniversity for professional and career development
- adaptation program and supervisor support at the start
- extended VHI (Voluntary Medical Insurance), preferential insurance for family and corporate pension program
- flexible discount on mortgage loans, equal to 1/3 of the CBR key rate
- Prime subscription with the ability to share with three close individuals
- reward for recommending friends to the Sber team.
