No items found.

Introducing Solo Enterprise for agentgateway

From pilots to production with context-aware AI networking

Solo’s vision for context-aware AI connectivity became reality with the introduction of the agentgateway project in April, becoming the industry’s first AI-native data plane to cover the full range of agentic use cases across LLMs, tools, inferencing, and multi-agent communication. The initial response was phenomenal and we realized that open participation and governance for the project would be just as important as it is for the protocols that it was supporting. We contributed agentgateway to the Linux Foundation in July with broad support from member organizations including Microsoft, T-mobile, Dell, CoreWeave, and Akamai for the project’s mission to build the future of agentic connectivity in an open, vendor-neutral community and governance framework. We couldn’t be happier with the incredible level of contribution and momentum the community has built, driving the project to over 1,000 GitHub stars in the first six months.

Throughout this journey, our mission has always been clear - help customers bridge the production gap and get AI and agentic applications to production safely, securely, and with governance controls required to leverage AI with confidence. Today, we’re excited to announce two initiatives to meet customers where they are and partner with them on AI connectivity:

  • Solo Enterprise for agentgateway : our production-ready enterprise distribution of the agentgateway project.
  • Solo Labs for MCP : a full-lifecycle engagement that pairs Solo’s forward-deployed experts with your team to design, build, and operationalize MCP tools from first use case to production. 

Solo Enterprise for agentgateway

Our enterprise distribution of agentgateway builds on the strong foundation of the open source agentgateway project, while adding the additional enterprise features, hardened distribution, and 24x7 support customers require to deliver AI connectivity to production. Solo Enterprise for agentgateway includes a broad range of enterprise-ready features based on our experience working with customers adopting agentgateway.

Full lifecycle MCP security and governance

Enterprises face new attack vectors as agents interpret natural-language context and invoke tools dynamically. Solo Enterprise for agentgateway goes beyond MCP authentication and authorization to support onboarding, registration, tool server fingerprinting, versioning, and runtime policy to protect against tool poisoning, rug-pulls, tool shadowing, and naming collisions before they start.

Last-mile security and identity

Enterprise deployments of MCP require more than just securing your MCP endpoint - authentication and authorization controls must extend to services tools are calling (e.g. Google, GitHub, Salesforce) for end-to-end security. Solo Enterprise for agentgateway provides an advanced implementation of secure elicitation to seamlessly integrate service identity with existing enterprise IdP and IAM infrastructure. Existing MCP tool servers that don’t support elicitation can be automatically wrapped allowing for immediate and uniform security without requiring additional development effort in your MCP server.

Verifiable least privilege security posture

Providing specific, enforceable, and auditable authorization controls is critical for controlling the surface area of attack vectors and hallucinations. Agentgateway supports automatic, secure token exchange for agent and tool interactions to isolate permissions to exactly what is required for an agent and tool to act on a user’s behalf. End-to-end audit trails for user interactions with agents and tools are cryptographically verifiable to support attestation and access reporting controls.

Drop-in integration for safe-by-default MCP deployment

Agentgateway drops into your existing agentic framework, IDE, or application as a virtualized MCP tool server requiring no refactors or code changes. Deploy as an Ambient waypoint for transparent traffic redirection, tool-server sandboxing, and automatic gateway injection. You can even import OpenAPI specs and expose them as governed MCP tools to accelerate adoption.

Beware of the Retrofit

It’s becoming increasingly common for incumbent vendors and existing API gateway projects to attempt to retrofit MCP support into their existing solution. This approach ends badly with real risks of prominent feature gaps, poor performance, and incomplete protocol support. Agentgateway was built from the ground up to support agentic protocols such as MCP, with native support for full protocol features, market leading performance, and the ability to virtualize and federate MCP tool servers to provide a single access and governance point for MCP interactions. When compared to an existing alternative AI gateway implementation, agentgateway offers 300x better memory utilization, 35x higher throughput, and more than 100x reduction in latency. 

Solo Labs for MCP

Solo Labs for MCP is a full-lifecycle, outcomes-driven engagement that pairs Solo’s forward-deployed experts with your team to design, implement, and operationalize MCP tools from first use case through production. Instead of months of trial-and-error POCs that stall before go-live, Solo Labs combines the engineering expertise behind agentgateway with your domain knowledge to accelerate adoption and de-risk deployment with security, observability, and production guardrails baked in.

Some of the key engagement points for the program include: 

  • Map and integrate your critical systems into MCP from day one, ensuring seamless tool connectivity.
  • Establish robust security, observability, and auditability practices tailored for MCP integration.
  • Co-develop any missing tool integrations to ensure your teams and agents have access to the enterprise data and systems they need.
  • Implement end-to-end observability and audit infrastructure so you can monitor every agent action with confidence.

Learn More and Get Involved

Cloud connectivity done right