Moscow, St. Petersburg Company: Bell Integrator

☑️Tasks: -Design and develop production systems using Python 3.12+ for AI agents and LLM models -Implement multi-agent architecture: integrate components, implement pipelines and custom orchestrators (LangChain, LangGraph, RAG) -Develop and support APIs (FastAPI, Django): create microservices, REST, asynchronous services -Integrate with relational, NoSQL, and vector databases (PostgreSQL, Redis, MongoDB, VectorDB: ChromaDB, Qdrant) -Create, train, and deploy ML models (ML pipeline, Prompt Engineering, RAG, quality assessment, A/B testing) -Containerization (Docker, Docker Compose) and deployment in Kubernetes, partial setup of DevOps processes (CI/CD: Jenkins, ArgoCD) -Develop unit tests, code reviews, performance optimization, security assurance -Collaborate in Agile teams, participate in planning, mentoring

☑️Requirements: -Proficiency in Python 3.12+: deep knowledge of standard and modern language features (asyncio, typing, dataclasses, pydantic, pytest) from 4-6 years of experience -Practical experience in developing AI/ML solutions using LangChain, LlamaIndex, RAG, integration with LLM API -Experience in designing and supporting high-load REST APIs on FastAPI/Django (architecture, DI, serialization, OpenAPI) -Advanced work with PostgreSQL/Redis, designing complex schemas, indexing, optimization, vector databases -Containerization skills, deployment in Kubernetes -Experience in implementing CI/CD pipelines, monitoring, alerting, logging organization -Solid knowledge of Git, experience in teamwork (Agile, Scrum) -English B2+ for reading documentation and communications

☑️Will be a plus -Fine-tuning and LLM customization, prompt engineering, quality assessment -Experience with MLOps, LLMOps, and multi-regional architectures -Participation in R&D or open-source projects, mentoring, architectural thinking -Working with GigaChat API

☑️Conditions: -Opportunity for professional and career growth within the company, chance to participate in various projects; -Experience working in a distributed team of professionals; -Salary level is discussed individually; -Work format: Office/hybrid in Moscow or St. Petersburg. Contacts: recruiter contact