Skip to main content
Short answers to the questions people ask most about Athena, with links to the full pages.

​
About Athena

​
What is Athena?

Athena is an enterprise AI analyst that works as a digital team member. It researches the web and your documents, analyzes data from connected databases, drafts reports and presentations, summarizes meetings, and automates recurring workflows. See the overview.

​
How is Athena different from a general-purpose chatbot?

Athena works inside your organization’s data and tools. It connects to systems such as SharePoint, Snowflake and Jira as each user, creates and edits real documents, sheets and presentations in a shared workspace, cites the exact source behind each answer, and runs repeatable workflows on schedules and triggers. Every action is attributed and can be undone.

​
What is the difference between Athena and Olympus?

Athena is the AI analyst. Olympus is the AI-native analytics and operations platform Athena runs on, which provides data integration, deployment options, APIs and applications. See Athena & Olympus.

​
Automation

​
What is an AOP?

An Agent Operating Procedure (AOP) is a reusable AI workflow. You write a prompt template with input parameters, such as text, file selectors or option lists, and run it on demand, on a schedule, by email, or through the API and SDK. See AOPs.

​
What is the difference between an AOP and a Flow?

An AOP is one reusable AI prompt with inputs, and the agent decides how to complete it. A Flow is a fixed sequence of steps you build in a visual, no-code editor. See Workflow automation compared.

​
How do I automate a recurring report?

Create an AOP that describes the report, then add a schedule trigger to run it daily, weekly or on chosen days. To start it from an email instead, add an email trigger. See Your First Workflow.

​
Security and data

​
Is Athena SOC 2 compliant?

Yes. Athena is SOC 2 Type II and HIPAA compliant. See Security and data handling.

​
Does Athena store or train on my data?

Athena does not train models on customer data, and its model providers operate under zero-data-retention agreements. Data from connected systems is fetched on demand as Live Assets: Athena keeps only a short-lived encrypted cache (one hour by default) and the identifiers needed to find the item again.

​
How does Athena handle permissions on connected data?

It depends on how the connection is set up. With a per-user connection, such as OAuth sign-in, each user connects their own account, Athena acts as that person, and the source system’s permissions apply to every request; if access is removed in the source system, it is removed in Athena immediately. With a shared connection, such as a BigQuery service account or an admin-registered SharePoint app, requests run with that account’s rights, so choose per-user connections where each person’s permissions must apply.

​
Can Athena be deployed in my own VPC or cloud?

Yes. Athena runs in its managed cloud, in your own cloud account on AWS (including GovCloud), Azure or GCP, on-premises, or fully air-gapped. See Security and data handling.

​
Does Athena support SSO and SCIM?

Yes. Athena supports SAML 2.0 and OIDC single sign-on with Microsoft Entra ID, Okta, Google Workspace, OneLogin, Ping Identity, ForgeRock and any compliant provider, plus SCIM for automatic provisioning and deprovisioning. See Single Sign-On (SSO).

​
Integrations and models

​
How do I connect SharePoint, Snowflake, BigQuery, Slack or Jira?

Each has its own setup guide: SharePoint, Snowflake, BigQuery, Slack and Jira. See All Integrations for the full list.

​
Which AI models does Athena support?

Athena supports 60+ models from multiple providers, all under zero-data-retention terms. You pick the model for a chat with the model selector, and you can set the model an agent or AOP uses in its configuration. See models. See Your First Chat.

​
How do I use the Athena API or SDK?

Install the Python or TypeScript SDK and authenticate with an API key. See the SDK Quickstart and the API overview.