Senior Data Engineer

Playrix; United States; Remote; USD 150,000 - 190,000; Full time

Job summary

Playrix is seeking a Data Platform Engineer to join the AI Platform team in the United States. You will own data pipelines and the platform layer used by AI products across the company.

You will work with modern data tools, profile production data, and help optimize LLM-related costs while building interactive services on top of data. This role offers mentorship and growth in a large engineering organization.

Qualifications

• Proven experience with modern data engineering stacks in production.
• Ability to design scalable data pipelines for AI platforms.
• Experience profiling, monitoring, and observability of data systems.
• Understanding of LLM usage, vector stores and RAG concepts is a plus.

Responsibilities

• Build and own data pipelines and the platform layer relied on by AI products.
• Analyze LLM usage, token costs, and AI spend reporting.
• Develop RAG systems including knowledge bases and vector/semantic search infrastructure.
• Ensure operability through profiling and monitoring of pipelines.
• Create interfaces and services on top of data with authentication and APIs.

Skills

Strong data engineering background

Backend mindset

Production profiling and monitoring

Tools

Databricks

dbt

Airflow

Dagster

Greenplum

Snowflake

Python

PostgreSQL

Redshift

Job description

About the role

AI Platform is one of our most forward-looking teams. It owns the shared infrastructure and platform layer that AI work across the company runs on - everything from how LLMs are used and measured to the knowledge and retrieval systems built on top of them.

A large share of the work is data engineering, because there's a lot of information to move, process and make sense of: LLM usage and cost analytics (tokens, cache, spend), and applied systems like contextual knowledge bases and vector search.

Playrix works with roughly 7 petabytes of data.

Responsibilities

• Building and owning data pipelines and the platform layer that AI products across the company depend on
• Analytics on LLM usage - cost, token and cache economics, financial reporting on AI spend
• RAG systems: contextual knowledge bases, vector and semantic search infrastructure
• Making the platform operable - profiling, monitoring, observability of the pipelines you own
• Building interfaces and services on top of the data, not just landing it (auth, APIs, interactive access for consumers)

Requirements

• Strong commercial data engineering experience with a modern stack - e.g. Databricks, dbt, Airflow, Dagster, Greenplum, Snowflake, or similar
• Confident Python, plus a working backend mindset: how authorisation works, how a service is exposed and consumed. We're looking for a data engineer who is comfortable building something interactive, not only delivering datasets downstream
• Practical understanding of how to profile and monitor data systems in production
• Real understanding of the AI/LLM layer - vector databases, embeddings, contextual and semantic search, RAG architectures, how autonomous agents work. This is an AI Platform team: we expect conceptual grasp of these technologies, not just having used a coding assistant
• Cloud experience (AWS ideally)

Our stack

PySpark, Databricks, dbt, Redshift, RDS, PostgreSQL

AI: Claude, KimiK3, and Codex with configured repository rules, a growing catalog of specialized skills, and agentic modes for specific tasks. MCP integrations with internal infrastructure (GitHub, Asana, Amplitude, etc.). Agentic workflows cover the entire cycle-from generating game elements and automated tests to automatically creating and reviewing PRs

We offer

The opportunity for continuous development in a team of 500+ professional engineers: we have a huge knowledge base and a mentoring system that allows you to adapt quickly.

Ability to move between areas and not only within development (Project Management, marketing, etc.).

Tasks that require you to make safe and effective architectural decisions as well as opportunities to apply interesting programming approaches.

An emphasis on developing each specialist's product ideas Time to play the games that you're developing, so you can envision how this or that feature will work for the user.

The ability to switch out processes and approaches for more efficient ones without lengthy approvals or bureaucracy.