Senior DevOps Engineer - AI Platform

We are building Manta to be a full-stack, trustable and auditable AI CMO agent.

This includes deploying state-of-the-art per-company infrastructure, distributed globally closest to the customer for optimal latency. By using an in-house Context Graph architecture the underlying storage requires hundreds of gigabytes of RAM per host to be able to reason on the underlying datasets, which demands cost-efficient deployments and extensive monitoring. By combining edge computing and dedicated server/on-prem hosting, we offer the best user experience and data privacy for our customers.

This creates unique engineering challenges around large-scale data processing, distributed systems, and high-performance infrastructure.

This Role

• Lead migration from managed hosting to a global Hetzner-based setup
• Implement Infrastructure as Code using Terraform or similar tools
• Build monitoring and observability across services
• Design deployment pipelines that support enterprise customers
• Contribute to building secure on-prem deployment capability
• Improve performance, cost efficiency, and system stability

What We’re Looking For

• Senior-level DevOps / Infrastructure experience
• Strong Terraform or similar IaC tooling
• Solid cloud infrastructure background
• Experience with Docker and containerized environments
• Monitoring setup experience such as Grafana or Prometheus
• Strong understanding of networking fundamentals
• Comfortable owning infrastructure end-to-end, along with developers
• You care about reproducible development environments and may already be using Nix, devenv, or similar tooling

Nice to have:

• Experience with bare-metal hosting
• Kubernetes/Dokploy/Docker Swarm experience
• Tailscale/Wireguard-based networking setups
• Startup or small-team experience
• Nix or other reproducible build systems

Why Join

• Real ownership of a post-LLM AI architectures’ infrastructure
• Work on enterprise-grade systems from early stage
• High impact in a small, senior team
• Modern stack without legacy baggage
• Remote-first environment