DevOps Engineer

About the role

Wanted: Colonist DevOps Engineer

Working Hours: We work asynchronously and value outcomes over hours

Location: Fully Remote

Compensation: Paid in USD. Negotiable based on experience and location

About Colonist

We're building the biggest digital board game platform on the internet. Players start 250k games every day. Last year: 60M games, 3,037 years of play. We're a fully remote, asynchronous team spread across multiple continents, focused on crafting extremely polished digital board games that people truly love.

About The Role

Most apps scale by splitting users apart. We can't. Players share one live game state. Every lag spike is felt mid-game. One viral video can 10x our traffic overnight. Your job: absorb 50k CCU without dropping a single game.

You will own the infrastructure behind every game session: clusters, deploys, databases, the edge, and the path from merge to production. You will make our infra knowledge shared, so no system depends on one person.

What You'll Do

- Own the stack behind every game: game, matchmaking and lobby servers on Kubernetes

- Manage cluster and cloud resources as code in Terraform

- Scale servers up and down with live traffic. Pay for players, not idle machines

- Ship updates often, without kicking players out of live games

- Keep terabytes of game data and replays fast and cheap. PostgreSQL, Redis, NATS under real load

- Defend the edge from bots and attacks. Cloudflare, rate limiting, Turnstile, R2.

- Own observability across Sentry and New Relic. Consolidate monitoring and close the gaps

- Document everything: systems, decisions, runbooks, so infrastructure knowledge lives in the team

- Build self-service paths so other engineers use infrastructure confidently.

- When something breaks, thousands notice in seconds. Find the root cause. Make it never recur

Requirements

- You love hard technical problems. Scale, latency and cost excite you

- You have senior-level experience in DevOps

- You have run Kubernetes in production: rollouts, autoscaling, networking, load balancers, and debugging a bad deploy

- You run infra lean. You optimize before adding hardware

- You document as you go. Runbooks and decisions are natural outputs of your work.

- You are comfortable with Linux, Docker, and a CI system you did not build

- You have operated PostgreSQL and Redis under real traffic: slow queries, connection limits, and schema migrations without downtime

- You have run a message broker in production; we use NATS, but equivalent experience transfers

- You have put Cloudflare, or a similar edge, in front of a public service

- You can debug a live incident across app logs, metrics, the database, and the edge