Keep the brain. Add the operating system.
Keep running your agent where it is. Add persistent identity, cloud-backed memory, integrations, channels, and one gateway without rebuilding the reasoning loop.
What your agent gains
The substrate your agent has been missing.
Keep the agent's brain. Drop the substrate work — identity, memory, channels, integrations, the model proxy, the gateway. All already running.
Persistent identity
A stable agent ID, scoped API keys, KMS-encrypted credentials. The agent's identity survives restarts, redeploys, and VM swaps.
Cloud-backed memory
Semantic vectors plus knowledge graph. Your agent recalls facts across sessions without you wiring a vector store.
20+ integrations
Notion, Xero, MYOB, M365, Atlassian, GitHub, Brave Search — install on the agent without writing connectors.
Every channel
Slack, Discord, Teams, Google Chat, voice / SMS / WhatsApp on a phone number, mobile and web. One agent, every surface.
Multi-LLM proxy
9 providers behind one endpoint with per-tenant policy. Switch the agent's model from the dashboard, no key rotation.
Agent-to-agent comms
Role-gated communication between your agents and any others in the org. Scoped sharing of memory, files, and integrations.
One command
npx @alfe.ai/cli setup
Run it on the box your agent runs on. The daemon installs, claims an identity, registers with the gateway, and hands you the substrate the moment it's online.
- →Works alongside your existing runtime — your agent's brain stays where it is
- →OAuth handoffs open in a browser — humans only sign the consent screens
- →Self-hosted or fully managed — same CLI, same outcome
How it works
Move in in three steps.
Install the CLI
Run `npx @alfe.ai/cli setup` on the box your agent runs on. It installs the daemon, claims an identity, and registers with the gateway.
Join the OS
The daemon comes online, the agent gets a persistent ID, and the substrate (memory, integrations, channels) becomes available.
Pick what you want
From the dashboard or via the agent self-service API, install integrations, claim channels, and grant scoped sharing — at any time.
Technical details
Built for the agents you've built.
Your agent doesn't move to Alfe. It joins the OS as a first-class citizen — keeping its own runtime, gaining the substrate it would otherwise have to reinvent.
Simple fleet economics
One plan. One pool. Every agent.
A tenant-wide credit pool funds model usage, voice, channels, storage, and compute. Add managed agents as the fleet grows.
Predictable at one agent. Coherent at one hundred.
Talk to Alfe
What should your agent be able to do?
Bring us the agent, the fleet, or just the workflow. We’ll help you choose the shortest path to a working system.
Architecture and platform questions
Managed agent builds
Enterprise and fleet deployments