Description
The Center for Practical Artificial Intelligence is engaged in the development and implementation of high-tech AI tools. Tasks are taken from the day-to-day business practice.
The first stage of selection for this vacancy is a conversation with an AI recruiter. After you apply, you will receive an invitation by email for an initial interview with GigaRecruiter on Telegram. The dialogue will take approximately 10 minutes. Its goal is to clarify missing details and expedite the review of your application.
GigaRecruiter is just starting its journey, so we ask for your understanding. Your experience and participation will help make it convenient and useful!
Responsibilities
- development and implementation of LLM agents and advanced RAG systems in the management domain to solve the bank's strategic tasks — from prototypes to production
- solving applied tasks for assessing the quality and completeness of provided documentation/reporting in business domains using AI assistants (RAG pipelines, query classification, answer generation, relevance assessment)
- research and configuration of multi-agent orchestration (LangGraph, LangChain, schema guided reasoning pipelines)
- working with GigaChat as the primary model, as well as experiments with ChatGPT, Gemini, Qwen
- fine-tuning models (instruction-tuning, adapters, LoRA, SFT, LLM-RL)
- development of quality metrics
- interaction with engineers and analysts — implementation of models into real use cases.
Requirements
- DS/ML work experience from 5 years
- deep expertise in NLP/LLM
- confident knowledge of development tools and infrastructure (bash, docker/openshift, git, etc.)
- fundamental knowledge in the field of ML, relevant education
- excellent Python language knowledge and experience in industrial development
- understanding of LLM architecture and prompt engineering principles
- experience in building RAG systems, fine-tuning and additional training of models
- ability to design inference / retraining pipelines
- experience in packaging models into services and interfaces (e.g., FastAPI, Flask, Tornado, StreamLit, ChainLit, etc.)
- experience in integrating LLMs with external APIs or knowledge bases.
Conditions
- comfortable modern office in Moscow, near Kutuzovskaya metro station
- office-based work format (hybrid format can be discussed after the probationary period)
- annual salary review and annual bonus
- corporate gym and relaxation areas
- more than 400 educational programs from SberUniversity for professional and career development
- extended voluntary health insurance, preferential insurance for family and corporate pension program
- flexible mortgage discount equal to 1/3 of the Central Bank's key rate
- free SberPrime+ subscription, discounts on products from partner companies.
