Platform Engineer
About the role
Stilta is building agentic AI for intellectual property. The Platform Engineer owns the platform Stilta agents run on, focusing on scalable, fast, secure infrastructure to support long-running, document-heavy workloads across millions of pages.
Responsibilities
- Design and run the Kubernetes platform behind large-scale agent swarms - scheduling, autoscaling and isolation across customers
- Build the systems that keep agent sessions fast and resilient under heavy, unpredictable load
- Own our infrastructure as code and the path to production on Azure, with enterprise-grade security, compliance and data residency built in rather than bolted on
- Make storage and data pipelines durable and performant for workloads spanning millions of documents
- Build observability that tells us why something is failing, not just that it is
Qualifications
- You've designed and scaled distributed systems in production, and you have well-earned opinions on what makes them reliable
- Deep expertise in Kubernetes, cloud infrastructure and infrastructure as code. We run on Azure with Terraform; equivalent depth elsewhere counts
- You're a software engineer first - strong in TypeScript and/or Python, and comfortable building the services that run the platform, not just configuring them
- You treat reliability, performance and cost as product decisions, and you instrument before you guess
- Bonus: experience running AI or inference-heavy workloads at scale, or with enterprise security and compliance
Skills
About Stilta
Agentic AI for intellectual property
Stilta builds agentic AI software for intellectual property, starting with patent litigation. Whether challenging or enforcing a patent, leading IP law firms and in-house IP teams use Stilta for source-backed, auditable analysis. Innovation drives the world's most important breakthroughs, but the tools that protect it haven't kept pace. Patent attorneys do high-stakes work with a process still built on manual document review and disconnected point solutions. Generative AI helps with drafting, but it doesn't deliver the trust and transparency that legal-grade work demands - no sources, no traceability, no audit trail. Stilta takes a different approach. We use agentic AI to run search and analysis at a scale manual methods can't match, with every output source-backed, referenced, and auditable. Until recently, this wasn't technically possible. We're four engineers from McKinsey/QuantumBlack who built and deployed AI systems at some of the world's largest organizations. With Stilta, we're equipping the people who protect the world's most valuable ideas with tools that match the stakes.
More open roles at Stilta
- AI Engineer5d agoHybrid (Stockholm, Stockholm County, SE) · Engineering