Reminix

Reminix is a platform for shipping AI services for products and internal teams. It supports agents, workflows, tools, and jobs as callable services, with projects described as private behind API keys and stateful by default. The service is built around ordinary HTTP routes, so code can be deployed and then called from an app or internal tooling.

The platform separates two common use cases. For product AI, it is intended for features that an app can call without building a backend around each one, with calls coming from an app, API routes, webhooks, or workers. For internal tools, it turns repeated internal work into private AI services that can be called from Slack, cron, admin tools, queues, or internal apps. The page names example uses such as support copilot, account research, document processing, Slack triage, ops runbooks, research jobs, and review queues.

Reminix says projects can be deployed from GitHub, and the example interface shows a connected GitHub repo, a service URL, model selection, scoped secrets, and live tail logs. It states that writing to /data persists without separate storage to provision, and that deploys, logs, environment variables, and secrets are already wired. It also says API keys are enforced at the edge before requests reach user code, and that no Reminix SDK is required because it uses plain HTTP. The supported stack names Hono, Express, FastAPI, LangGraph, Mastra, OpenAI, Anthropic, Gemini, and custom gateways, with the note that if something starts an HTTP server, it runs on Reminix.

The site includes sections for product AI, internal tools, templates, docs, pricing, and log in, but it does not provide pricing or licensing terms in the excerpted text. The page presents Reminix as a platform rather than a single app, and its own wording focuses on stable URLs, API keys, env vars, logs, deploys, and persistent state.