Senior Software Engineer, AI Gateway
Senior backend engineer on Docker's AI Gateway team, building the secure control plane that routes, observes, and governs AI agents and LLM workflows. Day to day: design Go/Rust services for model routing, credential handling, caching, rate limiting, and developer-facing OpenAI-compatible APIs.
- Compensation
- $160.9K–$260.7K / yr
- Location
- San Francisco, CA or Seattle, WA
- Work setup
- Remote
- Seniority
- Senior
- Employment
- Full time
- Experience
- 6+ years
- Category
- Software engineering
- Posted
Skills
Must have
- 6+ years of backend engineering experience with production-grade systems
- Deep knowledge of distributed and highly scalable systems, cloud-native infrastructure, and API design
- Experience building secure, high-throughput services (gateways, proxies, load balancers, policy engines)
- Fluency in Go, and/or Rust (both preferred)
- Familiarity with AI/ML platforms or model serving infrastructure
- Strong product mindset; excited about building developer-facing tools
- Ownership mentality with a bias for shipping, learning, and iterating
- Bachelor's degree in Computer Science, Engineering, or related field, or equivalent practical experience
Nice to have
- Prior experience with OpenAI, Anthropic, or similar LLM APIs
- Familiarity with RAG architectures, vector databases, or agent frameworks (e.g., LangChain, AutoGPT, CrewAI)
- Experience with policy engines (e.g., OPA), observability frameworks (e.g., OpenTelemetry), or API gateways
- Understanding of OAuth2.1, secret management, and cloud IAM systems
- Experience with Kubernetes, Docker, and microservices architecture
Full description
About Docker
Docker has been one of the most loved brands in developer tooling, trusted by more than 20 million monthly users and over 20 billion container image pulls. From solo founders to the world's largest companies, developers rely on Docker to build, share, and run their applications across our suite of products including Docker Desktop, Docker Hub, and Docker Scout.
We are a globally distributed, remote-first team building the tools that define how software gets built and delivered. As AI agents redefine software development, Docker is at the center of that shift, providing the sandboxed environments, verified images, and secure infrastructure that make autonomous workflows trustworthy by default.
_______________________________________________________________________
The Role
Docker AI Gateway is our answer to the complexity of taking AI agents from prototype to production. It’s a powerful, intelligent, and secure control point that eliminates the toil of model orchestration, tool management, observability, and governance—so developers can focus on building incredible AI agents, not gluing together infrastructure.
The Gateway sits at the center of modern AI applications, offering:
- A model and tool routing layer with built-in security and cost optimization
- A familiar OpenAI-compatible interface and MCP server
- Unified observability and policy enforcement
- Auto-RAG, tool injection, session summarization, and more
We’re just getting started—and we need exceptional engineers to help us build the backbone of the future of agent-based development.
_______________________________________________________________________
What You’ll Do
- Design and implement core systems powering the AI Gateway, including the model router, MCP gateway, and control plane
- Build infrastructure that supports dynamic model selection/routing, auto-failover, cost-based routing, and policy enforcement
- Own critical capabilities such as secure credential storage, session summarization, caching, and rate limiting
- Develop APIs for developers building with OpenAI-compatible interfaces and the Model Context Protocol
- Build the underlying infrastructure to support evaluation, telemetry, replay, and backtesting for agents and LLM workflows
- Lead architectural decisions and mentor engineers as the team scales
- Performance tuning - caching, session summarization, and rate limiting that keep the gateway fast under real production load