Run vs deploy
When your team is ready, you have two options: run it on Envelope, or deploy it to another platform. For most business users, running on Envelope is all you need. This guide explains the difference so you can make the right choice.
Running on Envelope
Running on Envelope is the default — and the simplest path. There's nothing to install, no infrastructure to set up, and no credentials to manage beyond what your agents need to do their work.
When you run a team on Envelope:
- The team runs on Envelope's runtime
- Results appear directly in your workspace room thread
- Run history is stored and searchable in your workspace
- You can trigger runs manually or on a schedule
This covers the vast majority of use cases. If you're running internal automations, data pipelines, report generation, outreach sequences, or anything else that produces outputs your team uses — Envelope is the right choice.
Scheduling runs
Once a team is running on Envelope, you can put it on autopilot. Schedules let your team run automatically on a recurring cadence — every hour, every day, every week, or a custom frequency — without you pressing a button each time.
See Running on autopilot for setup details.
Deploying to another platform
Some teams need to run on your own infrastructure — because of data residency requirements, because they integrate tightly with an existing system, or because you're already running AI workloads on a specific platform.
Envelope currently supports deploying teams to:
- Paperclip — self-hosted, on your own servers
- Amazon Bedrock — AWS-native AI runtime
- Relevance AI — no-code AI workflow platform
More platforms (including LangGraph, CrewAI, Vertex AI, and Azure AI) are coming soon.
When you choose to deploy to one of these platforms, Envelope packages your team definition and walks you through connecting your platform credentials. The team then runs on that platform's infrastructure rather than Envelope's.
Most business users never need this. Deploy to another platform only if you have a specific infrastructure requirement or are already deeply invested in one of these runtimes.
How to choose
| I want to… | Use | |---|---| | Get a team running as fast as possible | Envelope | | Run automations without managing servers | Envelope | | Schedule recurring runs | Envelope | | Keep data on my own infrastructure | Paperclip (self-hosted) | | Run on my existing AWS setup | Amazon Bedrock | | Use an existing no-code AI platform | Relevance AI |
Platform documentation
If you're deploying to a specific platform, see the setup guide for details: