RocketHuntSmart job search
МТС
Russia3 days ago
up to 250,000 RUB
per month
Reach out directly about this role
Grade
Middle
Work Format
Remote
Employment
Full-time
By company and country
AI writes it from your resume. You only have to send it.
Vacancy ID: MLm_DNA Rate: up to 250,000 RUB net per month. Grade: Middle+ / Senior. Load: Full-time. Format: Remote. Location/Citizenship: Russian Federation. Client: MTS.
About the project: building and developing modern MLOps platforms for machine learning with a focus on LLM solutions, RAG pipelines, and multi-agent systems.
We are looking for an MLOps engineer who is at the intersection of classic DevOps and advanced AI technologies. This role is for a specialist who not only knows how to build reliable CI/CD and orchestrate pipelines in Kubernetes but also deeply understands the specifics of deploying and maintaining LLMs, LangGraph, and Multi-Agent systems.
🚀 Key tasks: — Building, developing, and supporting MLOps platforms for machine learning. — Orchestration of ML pipelines and data processing (Apache Airflow). — Implementation and support of tools for experiment tracking, data and model versioning (MLflow, ClearML, Kubeflow, DVC). — Setting up and supporting CI/CD processes specifically for ML models and services. — Infrastructure support for the development and deployment of Multi-Agent Systems, RAG pipelines, and LLM applications. — Setting up monitoring and observability for ML services and LLM inference.
🎯 Mandatory requirements (Hard Skills):
MLOps and DevOps (CRITICAL!): — Deep understanding and experience with Docker and Kubernetes (K8s). — Experience in building CI/CD for ML (specific pipelines for training and deploying models). — Proficiency in MLOps tools: MLflow, ClearML, Kubeflow, DVC.
LLM and Agentic AI (CRITICAL!): — Understanding and experience with frameworks: LangChain, LangGraph, LlamaIndex. — Experience supporting infrastructure for Multi-Agent Systems and Autonomous Agents. — Understanding of RAG (Retrieval-Augmented Generation), MCP (Model Context Protocol), Prompt Engineering, and Tool Use principles. — Experience working with LLM provider APIs (OpenAI and similar).
Orchestration and Observability: — Solid experience with Apache Airflow. — Experience setting up monitoring and observability for ML systems (model quality metrics, data drift, latency, throughput).
💡 Will be a plus: — Experience optimizing LLM inference (vLLM, TensorRT-LLM, Ollama). — Python knowledge sufficient for writing automation scripts and custom Airflow operators. — Understanding of vector databases (Qdrant, Milvus, pgvector) and their operation.
