devops engineer for machine learning platforms

HireHi; Polska; On-site; PLN 180,000 - 300,000; Full time

Benefits offered by this job

Flexible work schedule

Medical insurance

Financial assistance

Job summary

Innowise ищет специалиста по MLOps-инфраструктуре для проектирования, развёртывания и поддержки устойчивых систем для ML-платформ и дата-пайплайнов. Работа с Python, SQL и облачными технологиями обеспечит надежность и масштабируемость ваших сервисов.

Кандидат должен владеть Docker, Kubernetes, GitLab/GitHub/Jenkins, MLflow/Kubeflow/ClearML и методами IaC; приветствуется опыт работы с Databricks и аналитикой данных. Гибкий график, работа с клиентами по всему миру, медицинское страхование.

Qualifications

• Опыт управления серверной или облачной инфраструктурой.
• Опыт работы с инструментами IaC; Docker и Kubernetes в продакшн.
• Опыт проектирования и поддержки CI/CD в продакшне.
• Опыт развёртывания ML-платформ (MLflow/Kubeflow/ClearML) и сопутствующих сервисов.
• Сильная Linux-система инжиниринга; Python и Bash для автоматизации.
• Опыт работы с Databricks и PostgreSQL, NFS, S3-совместимыми хранилищами.

Responsibilities

• Проектирование, разворачивание и поддержка инфраструктуры для ML-платформ и данных.
• Реализация Application Lifecycle Management для ML, автоматизация обучения, версионирования и развёртывания.
• Обеспечение надежности, масштабируемости и HA инфраструктуры MLOps.
• Проектирование распределённых вычислительных сред на bare metal, VM и облаке.
• Контейнеризация ML-сервисов Docker и Kubernetes; оркестрация релизов.
• Автоматизация Provisioning, lifecycle кластера и конфигураций via IaC.
• Построение и поддержка CI/CD через GitLab CI, GitHub Actions и Jenkins.
• Обеспечение наблюдаемости через логи, метрики и дашборды; RCA по инцидентам.

Skills

Python

Bash

Linux

Docker

Kubernetes

CI/CD

SQL

Databricks

MLflow

Kubeflow

ClearML

Observability

Go/Rust scripting

Tools

Terraform

Ansible

GitLab CI

GitHub Actions

Jenkins

Databricks

Kubeflow

Job description

Описание:

Innowise delivers technology projects for clients worldwide and has implemented more than 1600 cases for clients from the USA, Denmark, Germany, and other countries.

Задачи:

Design, deploy, and support resilient infrastructure for machine learning platforms and data pipelines using Python and SQL; Implement Application Lifecycle Management for machine learning, automating training, versioning, and deployment processes; Ensure reliability, scalability, and high availability of MLOps infrastructure and backend services; Design and manage distributed compute environments across bare metal, virtual machines, and private or public cloud; Containerize ML services and applications using Docker and Kubernetes, orchestrating production rollouts; Automate infrastructure provisioning, cluster lifecycle, and configuration using Infrastructure as Code; Build, integrate, and maintain CI/CD pipelines with GitLab CI, GitHub Actions, and Jenkins; Implement observability through logging, metrics, and dashboards for overall cluster health; Diagnose bottlenecks, resolve node and network failures, and conduct Root Cause Analysis as part of proactive incident management.

Требования:

Proven experience managing server or cloud infrastructure, Infrastructure as Code tools, Docker, and Kubernetes for at least 250 person-days; At least 300 person-days designing and maintaining CI/CD solutions in production environments; At least 200 person-days deploying and using MLflow, Kubeflow, ClearML, or similar platforms; At least 150 person-days in root cause analysis and stabilizing critical systems; Strong Linux systems engineering background with RHEL, Rocky, or SLES; Proficiency in Python and Bash for automation; Experience with PostgreSQL, NFS, and S3-compatible object stores; Experience with Data Analytics and Data Analysis using Databricks; Strong analytical mindset and an implementation-oriented approach; Ability to translate business requirements into scalable technical solutions; Excellent cross-functional collaboration with Data Scientists, Developers, and IT operations; Nice to have: deep experience with HPC schedulers and integrations, bridging HPC schedulers with cloud-native platforms and configuring dynamic scaling, Go or Rust scripting, SQL and PowerShell, MPI workloads and GPU scheduling, parallel file systems, configuration management with Ansible or Puppet.

Условия:

• Flexible work schedule;


• Experience working with clients worldwide;


• Financial assistance;


• Medical insurance.