Python Developer

Description

Toloka AI’s Machine Learning and Data Platform (MLDP) is, without exaggeration, a key part of Toloka’s success. The platform can be represented as a suite of tools and services designed to support the entire machine learning lifecycle, from data preprocessing to model deployment, management, and serving.

We are looking for an experienced Python developer (with a background of ML and Dev-/MLops as a benefit) who can help us make our platform even better.

Company Intro

Toloka AI is a global tech company that offers end-to-end AI solutions based on the effective combination of Crowd+ML+Expert+Tech Platform. Our solutions are powered by the most diverse global crowd from 100+ countries speaking 40+ languages.

We bring together experts from all over the world to form our growing team of more than 200 people. Our team is globally distributed and works from several locations, including the USA, the Netherlands, Israel, Czech Republic, Serbia, and more. Working at Toloka unlocks the door to being able to work on global projects and our commitment to personal and professional growth.

Responsibilities

• Develop ML infrastructure, for example, backend for LLM inference and training (mostly GPU);
• Automate data and ml pipelines;
• Implement new open-source tools for our internal clients;
• Maintain the platform;
• Develop technical documentation;
• Gather requirements from our internal ML teams and offer optimal solution balancing different use cases
• Be the part highly skilled MLOps engineers;

Requirements

• 5+ years experience as Software Engineer (primary python);
• Experience in infrastructure or backend development.
• Interest in working on internal infrastructure and ML Ops
• Ability to design solutions and maintain ML infrastructure
• English language - B2 or higher

REQUIREMENTS +

• Understanding of instruments and practices for LLM training and inferencing
• Experience with ML infrastructure
• Familiar with popular DevOps tools: Kubernetes, Docker, Terraform.
• Knowledge of how to work with Spark (Databricks).
• Experience in working with Airflow (or there's ETL orchestration tools) and DBT.

TOOLS AND TECHNOLOGIES

• MLFlow, Huggingface, vLLM, databricks, airflow, dbt, montecarlo, tableau, Microsoft Azure, Airbyte, k8s, terraform and a lot of custom self-developed tools.

Benefits

• Personal and career development: we support employee’s ambition for professional development and encourage them to implement their ideas in their projects;
• Flexibility: we offer remote or hybrid employment. You will also design with your manager a workday that works best for you;
• Paid parental leave + Annual vacation policy;
• Transportation expenses reimbursement;
• Contributions to personal pension plan.