FAQ

Frequently asked questions

Can't find your answer? Talk to us.

About HOLOH

What is HOLOH Workspaces?

One workspace for your team's chat, calendar and meetings, with AI agents built in. Your organisation decides what the agents can do, and confidential AI keeps sensitive work unreadable to the model provider.

Who is it for?

Teams and organisations that want their conversations, meetings and AI in one place, under their own control, and especially those whose work is too sensitive for ordinary AI tools.

Do we have to leave Microsoft Teams?

No. HOLOH members can join Teams meetings once your Teams administrator allows external participants, so teams can move at their own pace.

What do I need to join a meeting?

A browser, a webcam and a microphone. Guests join from the meeting link without installing anything. Chrome gives the best results.

AI agents

What can the agents do?

Answer questions from your team's conversations and tell you where they found the answer, write documents and do research. You mention an agent in a channel, and it replies in the thread.

What is "your Holoh"?

Your personal agent. It sees only what you can see and answers only to you.

What are coworker agents?

Agents shared by a team. Workspace admins set them up from templates or with the organisation's own instructions, tools and connectors.

Who decides what an agent can access?

Your organisation. Admins choose the models, tools and connectors agents may use, and HOLOH checks every request on the server. An agent can never see more than the workspace it belongs to.

Confidential AI and security

What is confidential AI?

In a confidential workspace, what you send to an agent is encrypted before it reaches the AI model, and the model runs only on sealed hardware. Neither the model provider nor the infrastructure that runs the model can read your prompts or the answers.

Who can read what we send to an agent?

Members who can already see it in that workspace. In a confidential workspace, the model provider and the infrastructure that runs the model can't read it either.

What happens if the sealed hardware can't be checked?

The agent stops and says the service isn't available. It never falls back to an ordinary model.

Is our data used to train AI models?

No.

Where can I read the legal terms?

In our Terms, Privacy Policy and Data Processing Agreement, linked in the footer.

What is sealed hardware?

Servers whose processors keep the AI model's working memory encrypted and walled off, so even the people who run the machines can't look inside. HOLOH checks the hardware before every request.

See HOLOH with your own team

Tell us about your team and we'll show you around.

Bring your team into HOLOH