Red Hat logo

Principal Software Engineer - AI Agent Platform (Czech Republic)

Red Hat · Brno - Tech Park Brno - B, Czechia

Information Technology Software & Development Hybrid Full-time Posted 4 days ago

About this role

The Red Hat Developer Tools organization is looking for a Principal Software Engineer to join the Kaiden team. Kaiden is an open-source platform for running AI coding agents — Claude Code, Goose, Cursor, OpenCode — in governed, isolated sandboxes. Every agent session runs inside its own container (via Podman) or microVM (via libkrun), while platform engineers control which models, tools, kills, secrets, and MCP servers are permitted through workspace-level governance

Kaiden is built by the team behind Podman Desktop, with over 4 million downloads. The project is an Electron-based TypeScript application with a Svelte UI, structured as a pnpm monorepo with a plugin architecture. Extensions integrate AI model providers (Anthropic Claude, Google Gemini, Mistral, OpenAI-compatible endpoints, OpenShift AI, Ollama, RamaLama), Skills providers, MCP server registries, and RAG tooling

The codebase also spans other languages (Go, Rust) for container runtime and sandboxing infrastructure, and integrates with Kubernetes for remote agent execution. The project is Apache-2.0 licensed and developed in the open at [github.com/openkaiden/kaiden](https://github.com/openkaiden/kaiden). The team works as a distributed, hybrid group spanning multiple time zones

We build in the open, contribute upstream, and welcome community participation. What You Will Do Implement sandbox isolation backends.** Design and evolve the dual-backend execution model (Podman containers and libkrun microVMs) that isolates each AI agent session, ensuring security boundaries, fast startup times, and resource governance. Contributing to the upstream backend solution to add features required by users is also required

Lead the workspace governance system.** Define and build the configuration and policy layer that lets platform engineers control which models, MCP servers, skills, secrets, and context sources are permitted per workspace, enforcing organizational AI policies at the platform level. Extend the plugi

Apply for this role on Red Hat’s official careers site.

Similar jobs