Description
We are developing an ecosystem of AI assistants at Sber. Our product helps developers and analysts find answers to technical questions faster, work with documentation, and knowledge bases. Currently, we are scaling the system, implementing agent-based scenarios, and connecting new data sources. We are looking for a Senior System Analyst who will take on the analysis and design of complex end-to-end scenarios, integrations, and requirements management.
Responsibilities
Responsibilities:
- API and Integration Design: Development of architecture and specifications for REST APIs for interaction between the AI assistant and external/internal systems (data providers, Sber's corporate services).
- Working with Event-Driven Architecture: Designing data flows through Kafka. Analysis and specification of event queues to ensure asynchronous interaction between the assistant's microservices.
- Agent Architecture Design: Formalizing requirements for AI agent logic. Describing scenarios where the assistant delegates tasks to internal tools (functions) and returns the result to the user.
- LLM Pipelines and Prompt Engineering (Analytics): Analysis of LLM answer quality. Participation in developing prompt structure, designing RAG pipelines (retrieval + generation), specifying requirements for ranking and context processing.
- Search Engine Development: Gathering and analyzing requirements for search improvement (semantic, hybrid), working with search result quality metrics.
- Requirements Management: Interaction with customers and consumer teams, decomposing tasks into the development backlog (Agile/Scrum), maintaining documentation in Confluence, building data models (logical and physical).
- AI Model Integration: Analytical support for the process of embedding new LLM models into the assistant's framework.
Requirements
Requirements:
- System analysis experience: 4-5 years (senior level)
- REST API: Confident use of API design tools (Swagger/OpenAPI), understanding of idempotency, pagination, and versioning principles.
- Kafka / Message Brokers: Understanding of topologies, partitioning, delivery guarantees. Experience in writing ASD (Architectural Scheme Documents) for event buses.
- LLM and AI Agents: Experience in developing and deploying AI assistants or chatbots in a production environment. Understanding of LLM, embeddings, and vector database principles. Experience in designing Agent-based systems (ReAct, planners).
- Databases: Confident SQL (PostgreSQL), understanding of NoSQL solutions (e.g., vector databases, Redis).
- Tools: Jira, Confluence, draw.io / PlantUML.
- Soft Skills: Systems thinking, ability to advocate for architectural decisions, skills in working with distributed teams.
Will be a plus:
- Understanding of C#/.NET at a code-reading level (to synchronize with the backend team).
- Experience in writing prompts and testing hypotheses.
- Experience in migrating from monolith to microservices architecture.
Conditions
Conditions:
- Office in Danilovsky Fort Business Center (Nagatinskaya, Tulskaya metro stations)
- Salary and social support for employees
- Extended voluntary health insurance from the first day of work for employees and preferential health insurance for family members
- Free SberPrime+ subscription, discounts on partner companies' products
- Corporate pension program
- Corporate training at the company's expense
- Employee referral program: you can invite familiar professionals to the team and receive a reward of up to 100 thousand rubles
- Powerful hardware, additional monitors, and everything needed for productive work
- Agile work with the best in the IT industry: 2000 product teams and the opportunity for internal mobility.
