HaloHalo
HaloHALOBUSINESS / PLANNED

For teams that want the tool, not the infrastructure project.

Your company’s AI workspace, inside your boundary.

HaloBusiness is a planned per-customer environment for coding, research, and operational work — with Guard at the front door, Primary behind it, and your company’s work kept inside the boundary.

Planned product / not yet available
COMPANY / WORKSPACEPRIVATE
Repository contextcompany-owned
Coding assistantinside your boundary
Answerready for your team
GUARD → PRIVATE LINK → PRIMARY

The idea

Make AI useful without making your team learn AI infrastructure.

01

Your company workspace

A company-workspace AI frontend for coding, research, and operational work — designed around the way your team already works.

02

No external model processing

The planned HaloBusiness boundary keeps company coding work within the company’s own environment rather than sending it to an external model or shared training pipeline.

Some frontier providers say they “cannot rule out that de-identified data derived from usage of our products helped improve our models.” HaloBusiness is intended for teams unwilling to accept that ambiguity: your business’s intellectual property and internal operational expertise stay within your isolated workspace.

HaloBusiness supports CMEK / BYOK setups, and Halo operators do not have access to your secure boundaries. On setup, your environment is initiated inside an encrypted container and set to non-public. Access is available only through secure authentication, using a method of your choice through the HaloBusiness frontend.

03

Halo handles the complexity

Your team gets a usable AI workspace without becoming an infrastructure company. The environment is handed over to your IT team to operate inside your own boundary.

The boundary, made visible

De-identification is not confidentiality.

Some frontier providers say they “cannot rule out that de-identified data derived from usage of our products helped improve our models.” Removing a name does not remove the value of your code, research, strategy, or operational knowledge.

HaloBusiness is intended for teams unwilling to accept that ambiguity. Customer work is processed inside a customer-specific Guard and Primary environment, outside a shared frontier-model improvement pipeline.

PER-CUSTOMER ENVIRONMENTNO SHARED MODEL PATH
01Employeelocal identity
02Guardsafety boundary
03Primaryprivate inference
local audit log / optionalmanual export / customer initiatedno Halo-operated request hop after handoff

Planned operating model

Choose focused capacity or a fully dedicated workspace.

Both planned options are per-customer environments. The cost-effective model provides dedicated capacity from 8am–8pm local time, at around 60% of the price of continuous access. Outside that window, the GPU / Guard instances may be used by a company in the opposite timezone.

A 24-hour option with a completely dedicated workspace is also planned for teams that need continuous access and exclusive capacity.

Each workspace is encrypted with its own keys and isolated from other workspaces. Without the relevant key and authenticated access, one workspace cannot read another workspace’s data.

HaloBusiness is planned, not live. The final workspace, deployment, availability, and pricing details will be confirmed before launch.

A separate product from HaloChat

HaloChat is for consumers. HaloBusiness is for company workspaces.

HaloChat

Consumer AI assistant for everyday conversations and personal work. Minimum necessary context may be sent to an eligible model or search provider.

HaloBusiness

Planned per-customer AI workspace for coding and operational work, with Guard and Primary inside the customer’s environment.

Want to shape the first release?

Tell us what your team would run.

We are collecting workload, workspace, and privacy requirements before HaloBusiness is implemented.

Ask HaloBusinessChat