Build the sovereign AI stack with us.

We are engineers, researchers, and operators working to make sovereign AI the default for the world's most critical institutions.

Why Carcosa

We're an early-stage team. What we offer is ownership, scope, and a mission worth showing up for.

Early-stage equity in a company building infrastructure for the world's most critical institutions.

Ship sovereign AI into environments where reliability, security, and ownership are non-negotiable.

Small team, wide scope. You will own real surface area from day one and shape what Carcosa becomes.

We're growing across every function. Find the role that fits.

Distributed systems and ML infrastructure engineers who want to build the secure runtime for sovereign AI.

Researchers focused on retrieval, evaluation, and on-device reasoning for regulated environments.

Practitioners who can take Hyperion through FedRAMP High, IL5, and FIPS 140-3.

Engineers who can deploy Hyperion inside customer datacenters and edge environments.

The heart of the Silicon Prairie.

Omaha, NE

Based in Omaha, at the heart of the Silicon Prairie, where leadership, engineering, and deployment operations come together.

The principles that shape how we hire, build, and ship.

Customers must own their intelligence end-to-end. Every design choice protects data residency, model ownership, and operational independence.

We build for the most regulated environments — secure deployments, signed bundles, and zero external dependencies are the default, not an upgrade.

Sovereign doesn't mean slow. We engineer Hyperion to deliver frontier model performance on local, on-prem, and edge hardware.

We build with the conviction that customers should own their intelligence.