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 plugin and extension architecture.** Evolve the plugin system and TypeScript extension API to support new AI provider integrations, MCP server registries, maintaining backward compatibility as the surface area grows.
Technical leadership and mentorship across the team.** As a Principal Software Engineer, you will influence and mentor engineers accross the team with collaboration with the team technical leader.
Contribute to upstream open-source communities.** Participate in and drive contributions to upstream projects that Kaiden depends on, including Podman, libkrun, the MCP specification, devcontainers, and related container and AI tooling ecosystems.
Mentor engineers and foster technical excellence.** Coach distributed team members through code reviews, design discussions, and pairing sessions. Raise the bar on testing, CI/CD practices, and security standards across the project.
Requirements
7+ years of software engineering experience, with demonstrated technical leadership on complex, multi-component systems.
Strong proficiency in TypeScript and experience building Electron or Node.js desktop applications; familiarity with Svelte or other modern frontend frameworks.
Working knowledge of Go and experience with container runtimes, OCI specifications, or virtualization technologies (Podman, Docker, libkrun, or similar).
Deep understanding of container orchestration (Kubernetes or Red Hat OpenShift), including deploying and managing workloads, writing controllers or operators, and working with cluster APIs.
Experience designing and building plugin or extension architectures, dependency injection systems, or SDK/API surfaces consumed by third-party developers.
Demonstrated ability to work effectively in distributed, hybrid teams across multiple time zones, communicating asynchronously through written design documents, pull request reviews, and issue discussions.
Practical experience using AI-assisted development tools (Claude Code, GitHub Copilot, Cursor, or similar) to accelerate coding, testing, and documentation - and the judgment to review and validate AI-generated output critically.
Track record of contributing to or maintaining open-source projects, particularly in the container, Kubernetes, or AI tooling ecosystems.
Experience with the Model Context Protocol (MCP), AI agent frameworks (LangChain, CrewAI), or building integrations for LLM-based tooling.
Familiarity with microVM technologies (libkrun, Firecracker, Cloud Hypervisor) and hardware-level isolation for untrusted workloads.
Experience with devcontainer specifications, workspace configuration standards, or developer environment provisioning
Benefits
Vision insuranceRemote work options
Additional Information
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.
Red Hat · Brno - Tech Park Brno - B