Orchestrate AI agents without the infrastructure
Define multi-agent workflows in markdown, route them through MCP or REST, and add knowledge with built-in RAG. One appliance runs it all in your own AWS account, on your own Amazon Bedrock.
From $0.149 per instance-hour or $949 per year, plus your own AWS usage.
Everything you need to orchestrate agents
Multi-agent workflows, MCP server routing and knowledge retrieval, in one appliance that runs in your AWS account.
Agent Orchestrator
Multi-Agent Workflows in Markdown
Define multi-step AI agent workflows in markdown. The compiler turns them into deterministic execution plans, then the orchestration engine runs each step with the right model, tools, and context. Approval gates, fan-out loops that run their branches in parallel, scheduled runs, and live progress in the console — all from a simple DSL.
MCP Gateway Pro
MCP Server Aggregator & Proxy
Connect AI agents to any combination of MCP servers and REST APIs through a single endpoint. Provide an OpenAPI spec and any REST API becomes MCP tools instantly — no server code needed. Built-in auth, rate limiting, routing, and analytics.
RAG Engine
Turn-key Knowledge Base for AI Agents
Upload documents and images, and your AI agents can search across them semantically. A built-in ingestion pipeline handles parsing, chunking, embedding, and indexing — exposed as MCP tools through the gateway. Supports PDF, Word, HTML, Markdown, CSV, JSON, and images with OCR.
One-Click Appliances
View allAppXen AI Chat
Self-Hosted Team AI on Bedrock
Multi-user AI chat powered by Amazon Bedrock, in your own AWS account: auto-HTTPS, MongoDB, keyless Bedrock via an instance role. No per-seat fees.
AppXen AI Builder
Visual AI & Agent Builder
Drag-and-drop builder for chatbots, RAG, and agents on Amazon Bedrock, running in your account: keyless Bedrock, PostgreSQL, hardened Ubuntu base.
AppXen n8n
Workflow Automation
Self-hosted workflow automation, production-hardened: auto-HTTPS, PostgreSQL persistence, firewall and automatic security updates. No execution limits.
AppXen App Builder
Internal Tools & Dashboards
Self-hosted Retool alternative: build internal tools, admin panels, and dashboards over your own databases and APIs. Auto-HTTPS, hardened Ubuntu base. No per-builder seat fees.
AppXen Git
Self-Hosted Git Service
Own your repositories, issues, pull requests, and CI in your AWS account: PostgreSQL, Git-over-SSH, auto-HTTPS, hardened Ubuntu base. No per-seat fees.
AppXen NextCloud
Self-Hosted Files & Collaboration
Own your files, calendars, and office docs in your AWS account: PostgreSQL + Redis, auto-HTTPS, hardened Ubuntu base. Unlimited users.
AppXen WordPress
Production WooCommerce Stack
WordPress + WooCommerce with a Redis object cache, MariaDB, auto-HTTPS and a hardened Ubuntu base. A real store in one click.
Built for the agentic era
Define multi-step agent workflows in plain markdown, wire them to any MCP tool, and let the platform handle execution. Backed by a built-in MCP gateway and RAG knowledge base — it all runs in your AWS account, billed hourly or annually through AWS Marketplace, plus your own AWS usage.
Workflows in markdown
Write multi-step agent workflows as markdown. The orchestrator compiles them into executable DAGs.
MCP gateway
Route agents to any MCP server, REST API, or built-in extension through a single endpoint, with an API key.
RAG knowledge base
Ingest documents, search with embeddings, and give your agents long-term memory out of the box.
One console for all of it
One sign-in to the console on your own appliance. Users have an admin or member role.
Health panel
- Your AWS identity, every service on the appliance, and Amazon Bedrock access for each model, on the dashboard home page
Agent Orchestrator
- Workflows with an editor and an AI copilot
- Executions with step output, and approve or reject at an approval gate
- A gallery of example workflows
MCP Gateway
- Add MCP servers and REST APIs; enable built-in services and extensions
- A gallery of official servers, file storage, analytics and an API playground
RAG Engine
- Upload and manage documents as sources
- Search your knowledge base from the console
API keys
- Create keys for your agents and the CLI
- Copy-ready settings for Claude Code, Claude Desktop and Cursor
Settings
- Invite users and set their role
- See the effective model registry and Bedrock status, read-only
- Change your password
Workflows and knowledge, from the terminal
Run agent workflows, ingest documents, and search your knowledge base from the command line. The AppXen CLI runs on your own machine and talks to your own appliance.
$ curl -fsSL https://appxen.ai/install.sh | sh Ready to deploy AI agents?
Runs in your AWS account on your own Amazon Bedrock. From $0.149 per instance-hour or $949 per year, billed hourly or annually through AWS Marketplace, plus your own AWS usage. Your data stays in your AWS account.