Isolated environment
A dedicated container or virtual machine per agent, with isolation between customers.
A safe environment for your agent.
We are building isolated environments for running AI agents, with clear resource limits, permissions and a log of every action. In development.
(01) What we are building
What we're working on building.
These capabilities are in development and not yet available. We announce availability once isolation and metering tests pass.
Secrets vault
Keys are stored encrypted and handed to the agent within narrow limits.
Network policies
Control what the agent may connect to outside its environment.
Resource quotas
CPU, memory and time limits to avoid unexpected consumption.
Human approval
Sensitive commands wait for your decision before they run.
Action log
Everything the agent did is logged and open to review.
(02) Before launch
Our conditions before launch.
- 01
Strict isolation
Tenant-separation tests before the service opens.
- 02
Abuse prevention
Limits, restrictions and monitoring of operations and cost.
- 03
Accurate metering
Reliable usage metering and invoicing.
Want to be among the first testers? Register your interest.
Contact usDetails, availability and pricing — when you get in touch.
What do you need to run today?
Have a server? Start by auditing it. Starting something new? Tell us what to launch and we'll help you choose the next step.