On-premises sandboxes
Run task workspaces on your network. Enterprise includes a supplied AI workstation1 for the subscription term; we install and maintain it, and ownership remains with us.
Deployment
Run your team where you need it.
On-premises, cloud or mixed. Choose where tasks and models run and the controls around them. We configure, monitor and support your Enterprise deployment.
Choose where tasks run
One location, or a combination in the same team.
Remote hardware
Run our sandboxes on remote or private-cloud infrastructure. We scope capacity, region and isolation requirements with your team.
Cloud sandboxes
Use cloud execution for tasks that need scalable capacity. Mix local and remote coworkers in one team. This deployment option is separate from the planned self-service Pro tier.
Choose where models run
Sandbox location and model hosting are separate choices. A local workspace can still call an external model.
- On your hardware. Host open-weight models on premises, sized to your GPU memory and budget. Local inference avoids sending requests to a model API provider.
- On rented GPUs. Host larger models without buying hardware. The infrastructure provider still hosts the processing; this is not an on-premises deployment.
- Through model APIs. Use approved providers for suitable work. Customers pay model usage directly; existing subscriptions depend on integration support and provider terms.
Mixed. Combine the approaches above: smaller local models for suitable document ingestion, stronger models for reasoning. This can reduce cloud-token usage; savings depend on the workload.
Model licences, connected tools, logging and network configuration still affect privacy. Self-hosted inference alone does not keep every part of a workflow on premises.
Enterprise
Setup and ongoing engineering
Setup covers installation, integrations and your initial coworker team. The monthly fee covers platform capacity, monitoring and a defined allowance of hands-on engineering.
Your forward-deployed engineer works on your business problems, not only on the platform: shaping workflows, building integrations, advising on how people and digital teammates should work together and working through delivery issues. The allowance is scoped to your engagement, not an exclusive full-time engineer.
Model-provider charges are separate. Cloud-only Pro plans are planned and not yet available. Contact us for a tailored Enterprise proposal.
Agree controls before onboarding
Access and credentials
Agree coworker permissions, approved tools, service accounts and credential management with your security team. Sensitive actions can require human approval.
Isolation and residency
Keep customer workspaces and operational context separate. Scope hosting regions, network access and storage around your data classification.
Retention, export and audit records
Journals, wikis and task records retain working context. Export them as portable packages at any time; agree retention, deletion and log access during onboarding.
Provider boundaries
Assess external models, fallback providers and connected tools under your data-handling policies. Your data is not pooled into another customer's knowledge base.
Compliance
Runs inside the environment you have already certified.
Enterprise is deployed on-premises, on a supplied workstation1 inside your estate. The system runs within your own security boundary, on your network, under your controls. Model hosting is a separate choice; see Choose where models run above. Your SOC 2, ISO 27001, HIPAA or GDPR programme can cover that environment in the same way it covers the other systems you run on-site.
We configure the deployment under your security policies and review processes. We scope the controls, the data flows and the supporting documentation your procurement and compliance teams require, and we answer what your auditor asks of us.
Digital coworkers can help gather audit evidence, check documentation against your chosen framework, flag gaps and draft responses.
What we do not claim. WeAreAIFirst holds no SOC 2, ISO 27001 or HIPAA certification, and we do not currently offer a Business Associate Agreement. We do not certify your deployment, and deploying on-site does not confer certification. Whether our software falls inside your certified scope is a decision for your own auditor. Your compliance team reviews the evidence and conclusions; coworkers do not replace an independent auditor.
1 Currently supplied: an NVIDIA GB10 Blackwell-based AI workstation. Hardware specification is agreed during setup.