The Center for Practical AI is a division of Sber that engages in complex AI projects at the intersection of applied research, engineering, and business implementation. We work on tasks where applying a ready-made model or building a quick prototype is insufficient: it requires understanding the subject area, testing approaches, designing solutions, and delivering measurable business results.

The team's current focus is on applied LLM agents, copilots, and agent systems for automating business processes. In such solutions, it's necessary to design the interaction between the agent and the user and internal systems, ensure correct tool selection, work with knowledge bases, consider domain limitations, and maintain a stable dialogue in text and voice channels.

We are looking for an Applied ML Engineer who will work at the intersection of ML/NLP, LLM engineering, and applied development. The main task is to create agent-based solutions for real business processes: design ML/LLM logic, test hypotheses, build working prototypes, establish quality assessment, and prepare solutions for pilot and implementation together with product and engineering teams.

The first stage of selection for this vacancy is an interview with an AI recruiter. After applying, expect a message from them in your messenger: the conversation will take approximately 10 minutes. The AI recruiter's task is to clarify any missing details about your experience and speed up the candidate review process.

Your responsibilities will include:

• Developing LLM agents, copilots, and components of agent systems for real business processes: testing hypotheses and bringing selected approaches to working prototypes.
• Implementing agent logic: processing user requests, selecting and calling tools, working with knowledge bases, and integrating with internal systems.
• Building development and quality assurance pipelines: defining necessary data and test scenarios, designing evaluations, conducting experiments, analyzing errors, and improving solutions based on results.
• Working at the intersection of research and engineering: reading papers, analyzing open-source solutions, quickly testing hypotheses, and bringing successful ideas to fruition.
• Identifying common patterns in specific business tasks and transforming successful solutions into reusable components and approaches.
• Interacting with clients, engineers, and product teams, clarifying requirements, and supporting solution piloting and implementation.

We expect from you:

• Practical experience with ML / NLP / LLM.
• Strong proficiency in Python and understanding of the ML stack.
• Experience developing LLM agents, RAG systems, or other LLM-based services, understanding modern agent architectures and approaches.
• Experience with LangChain or LangGraph, or independently implementing orchestration and tool-calling logic.
• Experience developing LLM solutions: from understanding data requirements and choosing architecture to evaluations, API/service development, and result demonstration.
• Experience developing the backend for ML/LLM services: integrations, containerization, basic CI/CD.
• Ability to quickly grasp new papers, frameworks, and approaches and turn new ideas into working prototypes.

It will be a plus:

• Experience independently managing ML/LLM services: from idea and prototype to implementation.
• Strong backend experience and the ability to build production-grade services.
• Experience working in startups or teams with a fast hypothesis testing cycle.

We offer:

• Comfortable modern office in Moscow, near Kutuzovskaya metro station.
• Option to choose a convenient schedule – office/hybrid (with at least 3 days a week in the office).
• Annual salary review and annual bonus.
• Corporate gym and recreation areas.
• Over 400 educational programs at SberUniversity for professional and career development.
• Extended voluntary medical insurance, preferential family insurance, and a corporate pension program.
• Flexible mortgage discount equal to 1/3 of the Central Bank's key rate.
• Prime subscription with the ability to share with three close contacts.
• Referral bonus for recommending friends to the Sber team.