MeowMeow: Gateway for modern AI workloads
What is MeowMeow?
MeowMeow is an AI gateway for MCP servers, AI agents, and custom backend workloads. You bring the workload — agent traffic, MCP tool calls, internal services, or APIs — and we build the gateway layer in front of it: routing, authorization, request handling, and protocol-aware modules configured around your actual traffic instead of a generic template.
The current release includes MCP, routing, static file serving, and authorization modules designed specifically for AI workloads and Model Context Protocol deployments.
What MeowMeow actually does
MeowMeow provides a single entry point in front of your services — agents, MCP servers, tool backends, and APIs — that parses and routes requests through a modular gateway architecture. Need authorization in front of an internal tool? Need to expose multiple MCP servers behind one endpoint? Need static asset delivery alongside API traffic? Add modules instead of rewriting infrastructure.
Who is it for?
If your AI workload needs a dedicated gateway layer built around MCP, agents, or custom tooling, you're in the right place.
AI Agent Platforms
Route and manage traffic across agents and tool calls without building the plumbing yourself.
MCP & Tool Builders
Expose and manage multiple MCP servers behind a single gateway while keeping authorization and routing centralized.
Infra & Platform Teams
Offload gateway development so your team can focus on product and infrastructure priorities.
Growing AI Workloads
Designed for systems with evolving traffic patterns and infrastructure requirements. Extend gateway behavior through modules rather than rebuilding your stack.
How it works
Most engagements ship within two to three weeks of the first conversation, depending on protocols, deployment requirements, and custom module work. We scope and price before any commitment — if something isn't ready yet, we'll tell you before you build around it.
Roadmap: meet the Kitties
MeowMeow's long-term scaling model is built around Kitties: lightweight worker instances that can be provisioned dynamically through platforms such as Kubernetes, Nomad, or AWS.
The current release focuses on gateway functionality: MCP handling, request routing, authorization, and static content delivery. Dynamic worker provisioning remains an active development target and is not part of the current release.
The goal is simple: keep the gateway focused on parsing and routing requests while allowing workload-specific processing capacity to scale independently when needed.
Technical profile
MeowMeow's gateway core is written in Erlang, chosen for the same reasons it has powered large-scale telecommunications systems: lightweight processes, fault isolation, and the ability to keep serving traffic while individual requests fail and recover cleanly.
Request handling is modular. The current release includes MCP, routing, static file serving, and authorization modules.
A custom context rule system allows request processing behavior to be composed from independent modules and applied selectively based on request characteristics and deployment requirements.
HTTP/1.1 is supported today. TLS termination, WebSocket support, and additional protocol modules are under active development.
The Kitties autoscaling architecture remains on the roadmap and is not part of the current release.
This isn't a drop-in SaaS product. Every engagement is a build: modules, routing behavior, authorization rules, and deployment strategy are configured around your workload rather than forced into a predefined template.
Frequently Asked Questions
What is an MCP gateway?
An MCP gateway sits in front of one or more Model Context Protocol servers and handles routing, authorization, request processing, and operational concerns through a unified entry point.
Can MeowMeow route multiple MCP servers?
Yes. MeowMeow is designed to expose and manage multiple MCP servers behind a single gateway endpoint.
What modules are currently available?
The current release includes MCP, routing, authorization, and static file serving modules. TLS termination, WebSocket support, and autoscaling workers are planned but not yet released.
How is MeowMeow different from a traditional API gateway?
Traditional API gateways are typically optimized around REST services and generic application traffic. MeowMeow is designed specifically around AI agents, MCP traffic, tool-calling workflows, and custom backend architectures.
Pricing
Per-engagement, no seats and no platform fees. Pricing reflects the modules, deployment model, traffic expectations, and custom work required for your workload.
Author
Flora
Rust-loving, Python-purring Rubyist with a taste for clean UI and warm naps in the sun.