Senior AI Engineer, Engine Reliability

[https://app.ashbyhq.com/api/images/user-content/1246dead-d30e-4c70-a1d3-ef7e5023d542/bad250f8-206a-41f4-9698-287303b69ce3/keyart%20supercell.png]

OUR ENGINE, TITAN, POWERS SOME OF THE HIGHEST-GROSSING GAMES IN MOBILE – TITLES THAT HAVE GENERATED BILLIONS IN REVENUE AND REACH HUNDREDS OF MILLIONS OF PLAYERS.

WE ARE INVESTING HEAVILY IN THE FUTURE OF TITAN, AND AS PART OF THAT INVESTMENT, WE'RE BUILDING A BRAND-NEW RELIABILITY TEAM: AN AUTONOMOUS, HIGH-SKILL GROUP THAT WILL WORK CLOSELY WITH OUR ENGINE AND GAME TEAMS, EMPOWERING THEM TO WORK SAFER, FASTER, AND EASIER THROUGH BETTER INFRASTRUCTURE, AI-DRIVEN ENGINEERING CAPABILITIES, AND FEEDBACK LOOPS.

AI is fundamentally changing how software is built, and we believe it will transform how games are made at Supercell. As AI agents take on more of the development process, the challenge isn't simply moving faster – it's building the confidence to do so. We're creating an AI-native Reliability team to help make that future possible, rethinking how we develop, validate, debug, and understand the impact of changes across real games. Our ambition goes far beyond better testing and automation: we're building the foundations for a world where engineers and AI agents can experiment, develop, and ship with a level of speed and confidence that simply isn't possible today.

A big part of how we work is through deep partnerships with our game teams. We don't want to build technology in isolation and hand it over for others to adopt. We work alongside the teams building and operating our games, understanding their challenges firsthand, experimenting together, and developing solutions that genuinely change how they work. We want to build capabilities that scale across Supercell while giving individual game teams the freedom to adapt and extend them to their own needs.

This role sits at the intersection of applied AI, game development, and engine programming. The challenge is the distance between a demo and a system that engineers trust: the evals that tell you it's actually working, the cost and latency that make it viable, and the failure modes you find before anyone else does. We're not looking for someone who wants to train a frontier model. We're looking for someone who gets unreasonable results out of the ones that exist, and knows how to turn emerging AI capabilities into real engineering value.

Did you know we have our own game engine? This article https://supercell.com/en/news/game-engine-called-titan/ explains why you've probably never heard of Titan, while this https://supercell.com/en/news/titan-game-engine/ more technical article is about making Brawl Stars toilet break-friendly.

WHAT YOU'LL BE DOING

- Design, build, and ship AI-native engineering capabilities that change how our engine and game teams develop, integrate, validate, and deliver software.

- Work directly with game teams, sometimes embedding with them, to understand their workflows, identify opportunities, prototype solutions, and take them all the way into real-world use.

- Build agentic systems that can take on increasingly complex engineering work, from understanding code changes and investigating failures to generating solutions, validating their behaviour, and providing actionable feedback.

- Own evaluation as a first-class engineering discipline. Build the evals and feedback loops that tell us whether our AI-driven systems are actually delivering reliable, useful results.

- Make autonomous agents reliable enough to trust with real work, with the right guardrails, human checkpoints, and observability.

- Build and evolve the underlying reliability infrastructure that makes AI-driven development possible, including integration validation, automated testing, and realistic environments across real devices and services.

- Think beyond individual tools. Help shape reusable capabilities that work across games while allowing teams to build and adapt their own solutions.

- Use judgment on where AI adds leverage, where deterministic approaches win, and where emerging technologies could fundamentally change how we work.

- Help set the practices of a new team as it grows, and raise the technical bar through code review, pairing, and design discussion.

WHAT YOU HAVE

- A track record of shipping LLM or agentic applications that real users used in production, not just prototypes.

- Strong, current command of modern generative methods at the applied level: you know what today's models are good at, where they break, and how to build reliably on top of them.

- Hands-on experience building agent workflows and tool use, and making them dependable through evals, guardrails, and observability.

- Real software engineering depth. You write code others can maintain, and you own things in production.

- Strong product and engineering instincts, with the ability to understand a team's problems, challenge assumptions, and turn emerging technology into practical solutions.

- Experience with build systems, CI pipelines, scripting, and developer tooling.

- Strong collaboration skills and comfort working closely with other engineering teams, navigating ambiguity, and driving work from early exploration through to adoption without adding unnecessary process overhead.

WOULD BE NICE IF YOU ALSO HAVE

- Games industry experience, and a real appreciation for mobile.

- Game engine or runtime development experience, and credibility with engine engineers.

- Professional experience with C++, enough to read and debug engine and runtime code.

- Experience with fine-tuning, model adaptation, or distillation, and the judgment to know when not to reach for it.

- Experience debugging cross-platform issues on iOS, Android, macOS, or Windows.

- Experience with device automation, device farms, or test frameworks such as Catch2 or gTest.

- Experience with observability tools such as Sentry, logs, metrics, or tracing, including applying AI to crash and log analysis.

- Exp…