How Firstweek handles your data.

Firstweek reads some of the most sensitive parts of a company: Slack, documents, tickets and the CRM. This page says exactly what it does with that access. If your review needs more, send your questionnaire to security@firstweek.app and we will answer it.

---

Connections

• Every connection uses the provider's OAuth and is read-only: Slack, Confluence, Jira, Google Drive, OneDrive and SharePoint, Microsoft Teams, GitHub, HubSpot and Salesforce.
• Firstweek only reads. Where a provider offers read-only permissions, Firstweek requests only those. GitHub (for private repositories) and Salesforce don't offer one, so their consent screens ask for broader access, and Firstweek's code never writes. In Salesforce you can enforce that by connecting an integration user with a read-only profile. In Slack, the Firstweek bot joins the public channels you select so it can read them.
• Your admin chooses which channels, spaces, repositories, folders and records are in scope. Nothing outside that scope is read.
• If someone else owns a tool, Firstweek sends them a request to approve the connection themselves.
• Any connection can be disconnected at any time from Settings → Integrations.

What is stored

• Raw messages and documents are processed in memory and never written to the database.
• What persists: AI-generated summaries, the onboarding artifact itself, and citations that link each claim to the thread, page or record it came from.
• Citations can show a short excerpt, up to 200 characters, cut from Firstweek's summary of the source, not from the raw message. This is on by default. A workspace admin can turn it off in Settings → AI → Show source references.

Encryption and isolation

• All traffic uses TLS. The database is encrypted at rest.
• Personal data and credentials (new-hire email addresses, SSO configuration, integration OAuth tokens) are also encrypted at the application level with AES-256.
• Every database query is scoped to your workspace at the application layer.

Access

• Single sign-on with Microsoft Entra ID or Google. A workspace can require SSO for everyone.
• Two-factor authentication for password sign-in.
• Role-based access: admins configure the workspace and connections, managers create and approve artifacts, new hires see only their own.
• Firstweek staff sign in with company Google accounts only. When staff view your workspace as one of your users, the start and end are written to your audit log. If you turn on impersonation approval, that user has to approve it first.

Audit log

Sign-ins, source connections and scope changes, artifact edits, approvals and deliveries, member and role changes, settings changes and staff access are recorded. Workspace admins can review the log in Settings → Activity log.

AI providers

• The content needed to write an artifact goes to Anthropic over TLS, with OpenAI as the fallback. OpenAI also turns the approved artifact into audio narration, and new hires' questions about their artifact go to the same providers.
• Email addresses, OAuth tokens, SSO credentials and payment data are never sent to AI providers.
• Neither provider trains on data sent through its API, and Firstweek does not use customer data to train models.
• Every provider is listed on the subprocessors page before it processes customer data.

Review before delivery

Nothing reaches a new hire until their hiring manager approves it. Managers can correct any statement before approving, and corrections are kept with the artifact.

Healthcare data

Firstweek is built for internal operational context, not patient data. It does not connect to EHR systems, and it does not sign BAAs today. Digital health pilots use only channels and spaces you confirm in writing contain no PHI.

Retention and deletion

• Disconnect any source, or delete any onboarding and its derived data, from the app.
• To delete a whole workspace, email legal@firstweek.app. We delete it within 30 days of your request, and backups are purged within a further 30 days.
• A Data Processing Agreement is available on request. See the privacy policy for the full terms.

Security reviews

Send the questionnaire to security@firstweek.app and we will answer it. For financial services, the 30-day pilot starts when your review clears, not before.

Questions reviewers ask

Does Firstweek store our Slack messages or documents?

No. Raw messages and documents are processed in memory and never stored. Only AI-generated summaries persist, with citations linking each claim to its source.

Are the connections read-only?

Yes. Every connection is read-only, scoped to the channels, spaces, repositories, folders and records your admin selects, and can be disconnected at any time.

Do you support SSO?

Yes. Your team can sign in with Microsoft Entra ID or Google, and a workspace can require SSO. Two-factor authentication is available for password sign-in.

Is there an audit log?

Yes. Sign-ins, source connections and scope changes, artifact edits, approvals and deliveries, member and role changes, settings changes and staff access are recorded, and workspace admins can review them.

Will you sign a BAA?

Not today. Firstweek is built for internal operational context, not patient data, and does not connect to EHR systems. Digital health pilots use only sources you confirm in writing contain no PHI.

Is our data used to train AI models?

No. Firstweek does not use customer data to train models, and its AI providers, Anthropic and OpenAI, do not train on data sent through their APIs.

---

Vulnerability disclosure

If you believe you have found a security vulnerability, report it to security@firstweek.app. Please include:

• A description of the vulnerability and the potential impact
• Step-by-step reproduction instructions
• Any relevant URLs, request/response payloads, or screenshots
• The type of issue (e.g. XSS, SQL injection, authentication bypass, data exposure)

We will acknowledge your report within 3 business days and provide an estimated timeline for resolution.

In scope

• firstweek.app and all subdomains: the web application, API, and authentication flows
• Data isolation between workspaces
• Authentication and authorization controls
• OAuth integration flows
• Exposure of encrypted credentials or tokens

Out of scope

• Denial of service attacks
• Social engineering or phishing of Firstweek staff
• Vulnerabilities in third-party services we depend on (report those directly to the vendor)
• Issues requiring physical access to our infrastructure
• Automated scanner output without proof of exploitability

Our commitments

• We will not pursue legal action against researchers who report in good faith under this policy
• We will not share your contact details with third parties without your permission
• We will credit researchers who responsibly disclose valid vulnerabilities, if they wish
• We ask that you give us reasonable time to remediate before any public disclosure