Why this role, why now
You join the Frontier Team, four people who act as Blockbrain's innovation hub. The Frontier is the edge of the map. We scout the AI landscape, prototype aggressively and validate opportunities before competitors do. That speed only matters if what we build survives enterprise load, security reviews and audits.
That is where you come in. You turn what our AI engineers build into production-ready systems and step in wherever architecture, scalability, security or operations become the real constraint. Nothing reaches production without your sign-off.
There is no playbook yet for running frontier AI in environments with the highest security requirements. That is why we are looking for a creative engineer. You treat constraints as design input, find new ways where standard patterns stop working and still have the judgment to say "not yet" when something isn't ready.
What we expect
- 7+ years in platform, infrastructure, SRE or DevSecOps roles with real exposure to AI or ML systems in production, plus a degree in Computer Science, Software Engineering, Electrical Engineering or a comparable field, or equivalent practical experience
- Production-grade expertise with Kubernetes, ArgoCD, Terraform, GitHub Actions, Prometheus and Grafana across at least two of AWS, GCP, Azure and on-prem
- Practical DevSecOps experience with threat modeling, secure SDLC, secrets management and supply-chain security, in environments shaped by ISO 27001, the EU AI Act or comparable standards
- Hands-on experience building secure, scalable environments for LLM workloads, including model serving, vector stores, RAG, observability and sandbox patterns
- Code literacy in TypeScript: you review backend code critically, push back where it matters and contribute targeted changes
- A creative problem solver: you enjoy problems without a known answer, find unconventional ways around technical and regulatory constraints and turn new ideas into patterns others can build on
- Pragmatic and self-directed: you ship effective solutions instead of collecting elegant designs, spot what needs doing without being told and explain trade-offs clearly, also in occasional technical conversations with enterprise clients
Expected AI Skills
- LLM systems understanding: you know how LLM architectures and inference pipelines work and where they fail in production. You don't train models from scratch, but you can judge what an AI system will actually do under real load
- AI infrastructure fluency: you are comfortable with model serving, RAG, embeddings, vector databases and evaluation tooling, and you have an informed opinion on what is worth adopting next
- AI as a daily multiplier: you use AI tools to improve DevOps workflows, automate infrastructure work and speed up your own delivery, try new tools early and feed what you learn back into the Blockbrain product
