At a glance
- Alfera is built for business workspaces. Your organization controls the workspace data submitted to Alfera.
- We process Slack, connected-app, file, browser, sandbox, billing, analytics, and AI-agent data only to provide, secure, support, bill, and improve the service.
- We do not sell Customer Data for money, and we do not use Slack Customer Data or Customer Data to train foundation models.
- You can revoke integrations at the source. Revocation stops new collection, but stored workspace data must be deleted through an authorized deletion request.
#Contents
- 1. Who this policy covers
- 2. Key definitions
- 3. Data we collect
- 4. How we use data
- 5. AI processing and model providers
- 6. Sharing and subprocessors
- 7. Security, storage, and transfer
- 8. Retention and deletion
- 9. Rights and choices
- 10. Cookies and attribution
- 11. Slack-specific commitments
- 12. Changes and contact
#1. Who this policy covers
This Privacy Policy explains how Alfera Inc. ("Alfera", "we", "us", or "our") collects, uses, shares, and protects information when you visit alfera.ai, use app.alfera.ai, install the Alfera Slack app, connect business tools, run AI agents, use scheduled tasks, book a demo, or contact us.
Alfera is designed for organizations. If you use Alfera through a company, team, Slack workspace, or other organization, that organization controls the workspace and the Customer Data submitted to it. Workspace administrators may manage members, connected apps, Slack access, billing, retention requests, and deletion requests for workspace-level data.
#2. Key definitions
| Term | Meaning |
|---|---|
| Customer Data | Data submitted to, generated in, or processed by Alfera on behalf of a customer workspace. This includes prompts, Slack content, connected-app data, uploaded files, generated artifacts, agent outputs, tool inputs and outputs, workspace memory, approvals, scheduled tasks, credentials, billing records, and service logs. |
| Personal Data | Information that identifies or can reasonably be linked to an individual, such as name, email address, account identifiers, workspace membership, IP address, support communications, usage records, or content that contains personal information. |
| Workspace administrator | A user authorized by the customer organization to manage workspace settings, members, billing, integrations, retention requests, or deletion requests. |
| Connected app | A third-party account, API, MCP server, custom HTTP endpoint, or automation provider that you authorize Alfera to use. |
#3. Data we collect
| Category | Examples | Why we collect it |
|---|---|---|
| Account, workspace, and authentication data | Name, email, avatar, user id, workspace id, workspace name, workspace role, organization membership, authentication provider identifiers, session state, and security metadata. | To authenticate users, provision workspaces, manage teams, enforce permissions, and secure accounts. Alfera uses WorkOS AuthKit and WorkOS widgets for sign-in, profile, organization, and team-management features. |
| Slack data | Slack workspace ids, installer/admin metadata, Slack user ids, names, email addresses if provided by Slack, channel ids, channel membership metadata, direct-message ids, events, thread ids, messages, files, and file metadata. | To operate the Alfera Slack app in conversations where Alfera is invited, mentioned, messaged, or configured to operate. |
| Conversations, files, artifacts, and memory | Chat sessions, prompts, model-visible context, assistant responses, run messages, tool calls, tool outputs, uploaded files, generated documents, spreadsheets, presentations, PDFs, screenshots, context summaries, inspector records, workspace facts, embeddings where configured, and keyword indexes. | To provide continuity, complete agent work, resume runs, debug failures, and make workspace-shared memory available to authorized users. |
| Connected-app credentials and tool data | OAuth setup state, access tokens, refresh tokens, expiration times, scopes, external account ids, account names, provider metadata, connection health, and tool execution records. | To set up, run, monitor, revoke, and debug connected tools. Stored credentials use an AES-256-GCM encrypted credential envelope when the production credential encryption key is configured. |
| Runtime, sandbox, and browser data | Sandbox ids, runtime environment ids, workspace mount keys, preview metadata, command inputs and outputs, logs, files, screenshots, downloads, browser outputs, generated artifacts, and temporary workspace state. | To run code, browser tasks, file operations, previews, and other agent work in local, Docker, or E2B remote sandbox environments. |
| Billing and payment data | Plan, subscription status, credit balance, credit grants, usage records, Stripe customer id, Stripe subscription id, checkout and portal metadata, invoices, payment-intent metadata, and tax or transaction metadata returned by Stripe. | To process subscriptions, credits, invoices, tax, usage, checkout, portal sessions, and billing support. Stripe handles full payment card numbers. |
| Website, analytics, attribution, and demo data | Pages viewed, product interactions, browser/device metadata, referrer, IP-derived metadata, error reports, analytics events, demo form fields, attribution identifiers such as gclid, gbraid, wbraid, and fbclid, and business-visitor identity signals resolved by third-party identification vendors where configured. | To operate the site, improve the product, debug errors, measure advertising, identify business visitors, and follow up on demo or purchase interest. |
| Communications with us | Name, email, company, role, message content, meeting information, diagnostic details, attachments, and security reports. | To respond to support, security, sales, privacy, and billing requests. |
Slack events we process
When the Alfera Slack app is installed, we may process app mentions, direct messages, public-channel messages where the bot is present, private-channel messages where the bot is invited, multi-party direct messages, app uninstall events, and token revocation events. Alfera verifies Slack event signatures when a Slack signing secret is configured and uses Slack OAuth scopes only to provide the Slack features you authorize.
Sensitive data and children
Alfera is not intended for children under 18 or the age of majority in their jurisdiction, and we do not knowingly collect child data. Alfera is also not designed for regulated sensitive-data workloads such as protected health information, payment-card storage, government classified information, or other high-risk regulated data unless we have a separate written agreement covering that use.
#4. How we use data
| Purpose | What that means in Alfera |
|---|---|
| Provide the service | Run AI agents, stream responses, execute tools, generate artifacts, build model context, maintain workspace memory, and resume or debug runs. |
| Manage access | Authenticate users, provision workspaces, manage teams, enforce workspace roles, and operate Slack and connected-app authorization. |
| Run workflows | Create, update, pause, delete, and execute scheduled tasks and trigger-based workflows. |
| Bill customers | Process subscriptions, credits, invoices, checkout sessions, customer portal sessions, payment webhooks, tax metadata, and usage records. |
| Communicate | Send service, security, billing, onboarding, and support communications through email or Slack where appropriate. |
| Improve and secure Alfera | Measure website and product usage, debug errors, monitor reliability, prevent fraud and abuse, improve product quality, and understand advertising attribution. |
| Comply and enforce | Comply with law, enforce our Terms, protect rights and safety, and respond to valid legal process. |
#5. AI processing and model providers
Alfera uses AI systems to generate responses, analyze context, run tools, draft documents, create summaries, classify information, search memory, and perform requested work. Relevant portions of Customer Data may be sent to AI providers only as needed to provide the requested service.
| AI data type | Examples |
|---|---|
| Inputs | Prompts, conversation history, files, images, tool schemas, workspace memory, and connected-app context. |
| Outputs | Assistant responses, summaries, classifications, tool-call arguments, generated artifacts, and draft content. |
| Operational metadata | Model name, provider route, usage metadata, request metadata, errors, and safety or abuse-monitoring metadata where applicable. |
Current model routes may include Anthropic, OpenAI, and Vercel AI Gateway depending on the configured environment, selected model, and availability. OpenAI-compatible APIs may also be used for embeddings and memory extraction where configured.
We do not use Customer Data to train our own foundation models. We do not knowingly opt Customer Data into third-party foundation model training. AI providers may temporarily retain API inputs and outputs according to their own API, enterprise, security, abuse-monitoring, and zero-retention terms. If you require stricter AI data-retention terms, contact us before submitting sensitive production data.
#6. Sharing and subprocessors
We do not sell Customer Data for money. We share information only as needed to provide, secure, support, bill, analyze, and improve Alfera, or as legally required.
Some analytics, advertising attribution, and conversion-measurement activities may be considered a sale, sharing, or targeted advertising under certain U.S. state privacy laws. You may contact us to exercise any applicable opt-out rights.
| Provider or category | Purpose | Data involved |
|---|---|---|
| Vercel | Hosting, previews, and serverless infrastructure for the marketing site and app. | Website, app, deployment, and operational data. |
| Google Cloud Platform | Cloud Run API hosting, Artifact Registry, Cloud Scheduler where configured, Secret Manager, logs, and related infrastructure. | API, logs, secrets, scheduling, and infrastructure data. |
| Managed PostgreSQL hosting | Database storage depending on the configured DATABASE_URL. | Users, workspaces, sessions, runs, messages, events, connections, billing state, memory, and artifacts. |
| WorkOS | Authentication, organization management, profile widgets, team widgets, and sessions. | Identity, account, workspace, and session data. |
| Slack | OAuth, messaging, event delivery, workspace communication, file upload/read features, and app installation management. | Slack workspace, user, channel, event, message, and file data. |
| Pipedream | Hosted connection setup, app catalog, proxy execution, and connected-app tool execution where configured. | Connected-app setup, credentials, scopes, provider metadata, and tool data. |
| Zapier | Offline ad conversion delivery to Google Ads and Meta, and optional automation or connected-app actions where configured. | Attribution, conversion, automation, and connected-app data. |
| Stripe | Checkout, subscriptions, invoices, customer portal, payment processing, payment webhooks, and billing metadata. | Billing, payment metadata, subscription, invoice, tax, and customer records. |
| PostHog | Product and website analytics where configured. | Usage events, user identification after sign-in, browser/device metadata, and analytics properties. |
| RB2B (Retention.com) | Website visitor identification and de-anonymization for B2B marketing where configured. | Website activity, IP-derived and device metadata, and identity signals used to resolve business-visitor identity, which may include a name, company, or professional email address. |
| Sentry | Error monitoring, performance traces where enabled, release diagnostics, and operational logs with default PII disabled. | Error, diagnostic, and operational telemetry. |
| Resend | Transactional and lifecycle email delivery where configured. | Email address, message metadata, and email content. |
| E2B | Remote sandbox compute, file operations, previews, browser tasks, and command execution. | Runtime files, commands, logs, screenshots, previews, and sandbox metadata. |
| OpenAI | Model inference, embeddings, memory extraction, or OpenAI-compatible API calls where configured. | Prompts, outputs, files, images, context, embeddings, and usage metadata. |
| Anthropic | Model inference for agent runs or related AI features where configured. | Prompts, outputs, files, images, context, and usage metadata. |
| Vercel AI Gateway | Model routing and API gateway services where configured. | AI request, response, routing, and usage metadata. |
| AWS | EventBridge Scheduler, relay Lambda, and related scheduling infrastructure where enabled. | Scheduled-task and infrastructure metadata. |
| Cal.com | Demo booking embed and meeting scheduling on the marketing site. | Booking, contact, company, and meeting metadata. |
| Meta Pixel and Meta Conversions API | Measurement on the marketing site and in the app for advertising audiences and reporting: page views, plus a demo booked, a signup, a first task finished and a paid plan. The same event can reach Meta from your browser and from our servers under one id, and Meta keeps one. | Pages viewed, browser/device metadata, IP address, Meta cookies such as _fbp and _fbc, and for conversions a hashed email address and a hashed user id. |
| Google Ads and Meta | Advertising attribution and offline conversion reporting where configured. | Ad click ids, landing URL, landing time, conversion metadata for demo bookings, purchases, and ad-attributed workspace signup and activation milestones (including contact email and name where available), and related attribution data. |
| Customer-authorized connected apps | Any third-party app, API, MCP server, or custom HTTP endpoint that you connect or ask Alfera to use. | Data and actions authorized by your scopes, account settings, and instructions. |
We may also disclose information when required by law, legal process, security investigation, fraud prevention, enforcement of our Terms, protection of rights and safety, or as part of a merger, acquisition, financing, corporate reorganization, bankruptcy, or sale of assets, subject to appropriate confidentiality protections.
#7. Security, storage, and transfer
We use administrative, technical, and organizational safeguards appropriate to the nature of the service.
| Safeguard | Current approach |
|---|---|
| Transport security | TLS in transit. |
| Storage security | Cloud-provider encryption at rest where supported. |
| Credentials | AES-256-GCM encrypted credential envelopes when the production credential encryption key is configured. |
| Access control | Least-privilege access controls, authentication controls, and workspace permissions. |
| Webhooks | Signature verification where supported and configured. |
| Observability | Audit logs, run events, service logs, and monitoring for reliability and abuse prevention. |
No system is perfectly secure. You are responsible for managing access inside your Slack workspace, source systems, and connected third-party accounts.
Production infrastructure is primarily hosted in the United States unless otherwise stated in a separate agreement. Customer Data may be processed in the United States and other regions where our subprocessors operate.
#8. Retention and deletion
| Data category | Typical retention approach |
|---|---|
| Workspace content and agent records | Retained while needed to provide Alfera, maintain workspace continuity, debug runs, satisfy contractual obligations, and support customer requests. |
| Connected-app credentials | Retained while the connection remains active or as needed to revoke, audit, or resolve issues. |
| Billing records | Retained as required for tax, accounting, payment, dispute, fraud prevention, and legal obligations. |
| Logs and diagnostics | Retained for reliability, security, fraud prevention, debugging, and operational needs. |
| Backups | May persist until backups expire in the ordinary course. |
If a workspace is closed or we receive a validated deletion request from an authorized administrator, we will delete or de-identify relevant Customer Data from active production systems within a commercially reasonable period, unless retention is required by law, security, fraud prevention, billing, dispute resolution, or legitimate business needs.
Revoking a connected app, uninstalling Slack, or disconnecting an integration stops new collection from that source but does not by itself delete previously stored Customer Data. You may contact us to request deletion.
#9. Rights and choices
Depending on your location, you may have rights to access, correct, delete, restrict, object to, or receive a copy of personal data, and to opt out of certain processing such as sale, sharing, targeted advertising, or certain profiling.
To exercise rights, contact hello@alfera.ai. We may need to verify your identity and authority, especially for workspace-level data controlled by an organization.
If you are in the EEA, UK, or another jurisdiction with similar rules, our legal bases may include performance of a contract, legitimate interests, consent, and compliance with legal obligations. Where required for international transfers, we use appropriate safeguards such as Standard Contractual Clauses or applicable data-transfer mechanisms.
#10. Cookies and attribution
We use cookies and similar technologies for sign-in, session security, app routing, product analytics, error diagnostics, and advertising attribution.
| Cookie or technology | Purpose |
|---|---|
| WorkOS session cookies | Authentication and session security. |
alfera_authed hint cookie | Lets the marketing site show app navigation when a user has signed in. |
| PostHog analytics where configured | Product and website analytics. |
| Sentry diagnostics where configured | Error diagnostics and operational monitoring. |
alfera_attribution first-party cookie | Stores ad click identifiers such as gclid, gbraid, wbraid, fbclid, the landing URL, and landing time for up to 90 days. |
Meta Pixel (_fbp, _fbc cookies) | Tells Meta which pages of the marketing site and the app you viewed, and which ad you came from, for advertising audiences and ad reporting. Never on sign-in, invitation, or app-connection pages. |
| RB2B (Retention.com) visitor identification where configured | Identifies business visitors to the marketing site and the app (never on sign-in, invitation, or app-connection pages) and may associate site activity with professional contact data such as a name, company, or email address. |
If you book a demo, purchase through Stripe, or sign up and use a workspace after arriving from an ad, attribution data may be passed through our API and Stripe metadata so we can report offline conversions (including signup and activation milestones) to Google Ads and Meta through Zapier where configured.
When you visit our website, cookies and similar technologies may be used by our online data partners or vendors, including RB2B (Retention.com), to associate these activities with other personal information they or others have about you, including by association with your email. We (or service providers on our behalf) may then send communications and marketing to these email addresses. You may opt out of receiving this advertising by visiting https://app.retention.com/optout. You may also opt out of the collection of your personal data in compliance with GDPR by visiting https://www.rb2b.com/rb2b-gdpr-opt-out.
On alfera.ai, visitors from Europe see a cookie banner. Until they say yes, the Meta Pixel, RB2B, and ad-click cookies do not run, and a demo booking reaches Meta from our servers only after a yes. The answer is kept in the alfera_consent cookie for 180 days. The app has no banner: creating an account accepts our Terms, which allow this measurement in the app. A "No thanks" given on alfera.ai still turns it off in the app, and you can email hello@alfera.ai to stop it at any time.
You can control the ads Meta shows you, and how it uses activity from other sites, at https://www.facebook.com/adpreferences.
You can control cookies through your browser settings. Some cookies are necessary for authentication, security, routing, or billing flows and disabling them may break parts of Alfera.
#11. Slack-specific commitments
- We use Slack data only to provide, secure, and support Alfera.
- We do not sell Slack data.
- We do not use Slack message content for advertising.
- We do not use Slack APIs to develop, improve, or train generalized AI or ML models.
- We do not train our own or third-party foundation models on Slack Customer Data.
#12. Changes and contact
We may update this Privacy Policy from time to time. If we make material changes, we will notify customers through reasonable means, such as posting the updated policy, emailing administrators, or notifying users in the app. The "Last updated" date above shows when this policy was last revised.
For privacy requests, support, or questions, contact hello@alfera.ai. For security reports, contact security@alfera.ai.