Privacy Policy.
This Privacy Policy explains how Deepslate Inc., doing business as Embrasure (“Embrasure,” “we,” “us,” or “our”), collects, uses, discloses, and protects personal information when people visit our website, evaluate or buy Embrasure, use the Embrasure product, or contact us. For customer deployments, this public policy is supplemented by our Data Processing Addendum (“DPA”) and any signed agreement with the customer.
Scope and roles
Embrasure provides software that helps teams connect governed data sources, ask questions, build context, and operate data agents. In most customer workspaces, the customer is the controller of the personal information they choose to put in or connect to Embrasure, and Embrasure is the processor or service provider. Embrasure is the controller for account administration, website visitors, sales, billing, security, and product improvement data we collect for our own business operations.
This policy applies to Embrasure websites, hosted product interfaces, local tooling that authenticates to Embrasure services, support communications, and related business interactions. It does not govern a customer's own data practices or third-party websites that may link to or from Embrasure.
Information we collect
| Account | Name, work email, company, title or role, workspace membership, authentication identifiers, SSO domain, invited users, and administrative settings. |
|---|---|
| Customer data | Prompts, agent instructions, dashboard content, notebook content, uploaded or connected metadata, semantic definitions, query plans, and rows returned to a chat, workflow, notebook, or dashboard session. |
| Connector data | Connection configuration, hostnames, schemas, table and column names, access scopes, sync status, and encrypted credentials or tokens provided by authorized workspace administrators. |
| Usage | Pages and features used, searches, runs, model selections, errors, latency, logs, device/browser information, IP address, approximate location derived from IP, and audit events. |
| Support | Messages, call notes, implementation details, bug reports, screenshots, and other information shared with our team. |
| Commercial | Plan, billing contact, invoice status, contract metadata, procurement information, and communications with sales or customer success. |
Embrasure does not intentionally collect payment card numbers directly; payment information is processed by our payment or billing providers if a paid plan requires it.
How we collect information
- Directly from users and workspace administrators when they create accounts, configure workspaces, connect data sources, invite teammates, contact us, or use the product.
- Automatically from the product and website through logs, cookies, analytics events, security telemetry, and similar technologies.
- From connected systems that an authorized customer user chooses to connect, such as warehouses, identity providers, collaboration tools, or business applications.
- From business sources such as procurement systems, partner referrals, public company websites, and communications with prospective customers.
How we use information
- Provide, secure, maintain, and improve Embrasure.
- Authenticate users, enforce workspace roles, support SSO and domain restrictions, and prevent unauthorized access.
- Connect to customer-authorized data sources, generate catalog metadata, run approved queries, produce agent outputs, and maintain auditability.
- Provide support, debug issues, investigate errors, and communicate about product changes, incidents, billing, and security.
- Measure usage, reliability, performance, and feature adoption so we can improve the product and prioritize engineering work.
- Protect against abuse, fraud, security threats, and violations of our agreements or acceptable use rules.
- Comply with legal, tax, accounting, security, and compliance obligations.
Google Workspace directory data
A workspace administrator may connect a Google Workspace directory using Google OAuth. Embrasure requests read-only access to directory users and groups. This can include names, work email addresses, account status, organizational information, group membership, and profile photos.
| Purpose | Sync the customer's company directory, show linked profile photos, grant Embrasure membership and roles from selected Google Groups, and deactivate access when a linked directory account is suspended or removed. |
|---|---|
| Storage | Directory records are stored within the customer's Embrasure workspace. OAuth tokens are encrypted at rest. A copy of a profile photo is stored only for linked, active users when photo sync is enabled. |
| Sharing | Google Workspace directory data is not sold, used for advertising, or shared except with service providers needed to securely host and operate Embrasure, when directed by the customer, or when legally required. |
| AI use | Google Workspace directory data is not used to train AI models and is not sent to model providers for prompts, enrichment, or generation. |
| Control | Workspace administrators choose which groups grant access and whether profile photos sync. Google access can be revoked from the connected Google Account or the Google Admin console. |
| Deletion | Google Workspace directory data is deleted or de-identified after workspace termination or a verified deletion request under the retention periods described below, subject to backup, security, and legal requirements. |
Embrasure's use and transfer of information received from Google APIs adheres to the Google API Services User Data Policy, including its Limited Use requirements.
Google Sheets data
A workspace administrator may connect a Google account and choose a spreadsheet through Google Picker. Embrasure requests access only to Google Drive files the user explicitly selects for the app. It does not request access to browse every file in the user's Drive.
| Purpose | Import the selected spreadsheet's tabs as governed warehouse tables and refresh them on the schedule configured by the customer. |
|---|---|
| Storage | Selected spreadsheet values and metadata are stored in the customer's Embrasure workspace and managed warehouse. OAuth tokens are encrypted at rest and refreshed for scheduled synchronization. |
| Sharing | Google Sheets data is not sold or used for advertising. It is shared only with service providers needed to securely host and operate Embrasure, when directed by the customer, or when legally required. |
| AI use | Google Sheets data is not used to train general-purpose AI models. Customer-authorized product features may process selected data under the workspace's configured controls. |
| Control | Administrators choose each spreadsheet Embrasure can access. They can remove the connector in Embrasure or revoke access from their Google Account. |
| Deletion | Removing a connector stops future synchronization. Imported data is deleted under the workspace's configured deletion and retention controls, subject to backup, security, and legal requirements. |
Embrasure's use and transfer of Google Sheets data adheres to the Google API Services User Data Policy, including its Limited Use requirements.
Google Cloud SQL onboarding
When an administrator connects Google during Cloud SQL setup, Embrasure reads instance and database metadata in the Google Cloud project they enter. This includes instance and database names, region, PostgreSQL version, connection addresses, network and SSL settings, and database flags. We use this information to help the administrator choose a database and check connection and sync prerequisites. We read your Google account identity to identify the connected account and protect shared access during cleanup. When you choose passwordless access, your verified Google account email also identifies the database user.
Google's Cloud SQL permission allows management of Cloud SQL resources. Embrasure uses it only for metadata reads during onboarding; it does not change instances, networks, database flags, or database users through the Google API. For Scheduled sync, administrators can use their Google account when IAM database authentication and read permissions are configured, or supply a PostgreSQL login. Continuous sync uses a PostgreSQL login with replication access.
| Storage | Google credentials are encrypted at rest. We request offline access only when you choose Google account login for passwordless Scheduled sync. Setup sessions expire after 30 minutes, and unused setup credentials are cleared on completion, expiration, or maintenance cleanup. Passwordless connectors retain encrypted refresh credentials for ongoing reads; connectors using a PostgreSQL login retain their separately encrypted database credentials. Unused Google grants are revoked by a retryable cleanup process and their retained credentials are then deleted. Shared grants remain available while another source or unexpired setup needs them. Legacy grants without an identified account are retained conservatively while another Google source remains active. |
|---|---|
| Sharing and AI | Cloud SQL data is not sold, used for advertising, or used to train general-purpose AI models. Data is processed by service providers needed to host, secure, and operate the features the customer uses. Google tokens are used for onboarding API calls and authorized IAM database login, not AI prompts. After a database connection is configured, customer-authorized features may process selected database data under the AI and workspace controls described in this policy. |
| Control | You can revoke Embrasure's Google access in your Google Account connections. This prevents future token refreshes for passwordless sources and may affect other Google connections in the same project. Existing database sessions can remain active until closed. Revoking Google access does not remove a separately configured PostgreSQL login. Remove the connector in Embrasure to stop future synchronization, and revoke the database login or its grants in PostgreSQL to remove database access. |
| Deletion | Removing a connector does not automatically erase data already imported into the managed warehouse. Connection metadata and imported data are subject to the retention and deletion terms below. Request deletion through your workspace administrator or privacy@embrasure.ai. |
Embrasure's use and transfer of information received from Google APIs adheres to the Google API Services User Data Policy, including its Limited Use requirements.
Ember for Slack
A workspace administrator may install Ember, Embrasure's Slack app. Ember responds when someone sends it a direct message or mentions it in a conversation it has been added to. It reads only conversations it can access in Slack, and only when needed to answer a request.
| Data accessed | The request and its thread, recent messages in conversations Ember has been added to when a request needs them, and the conversation and member information Slack provides to check who can see an answer. Ember does not request personal Slack search permissions. |
|---|---|
| Purpose | Answer requests, show the sources behind an answer, deliver scheduled reports and approved messages, and keep the conversation history needed to follow up in the same thread. |
| Access checks | Before Slack content is read, reused, or delivered, Ember checks that the person asking and everyone who will see the answer can currently see the source conversation. Edits and deletions in Slack replace or remove the stored copy used for later replies. |
| Storage | Requests, answers, and Slack identifiers are stored in the customer's Embrasure workspace and hidden from other product views. Slack tokens are encrypted at rest. |
| AI use | Requests and the Slack content needed to answer them are sent to the AI model providers listed on our subprocessor page. We request zero data retention and no provider data collection, and we do not train models on Slack data. |
| Uninstall | Removing Ember stops new access immediately. Slack content and data derived from it for that installation, including stored conversations, answers, findings, report setups, and the Slack tokens, are deleted within 14 business days, normally within two days. Usage and billing totals are kept without message content. |
| Other copies | Encrypted database backups expire within 30 days. Workflow history, which can include short progress notes, is kept for 7 days after a request finishes. Operational logs are kept for up to 365 days and can include error text. Agent traces sent before October 4, 2026 to Braintrust (September 3 to 11, 2026) and ProperMotion (from September 11, 2026) may contain request and answer text; they remain subject to those providers' retention and have not been confirmed deleted. Traces since then contain only technical timing and usage information. |
Warehouse and customer data
Embrasure supports direct queries to connected sources and ingestion into a managed warehouse. When a customer configures ingestion, Embrasure copies the selected source tables and synchronizes them using the configured sync mode. For direct queries, we may store the metadata, semantic definitions, query plans, prompts, agent activity, and bounded result rows needed to provide the product, maintain conversation history, support audit logs, and troubleshoot customer-reported issues.
Customers control the credentials, scopes, workspace roles, governance settings, retention expectations, and connected sources they configure. Customers should not connect data sources or provide personal information unless they are authorized to do so.
AI processing and model training
Embrasure uses large language models and related AI services to generate responses, plans, summaries, workflows, and other product outputs. Input to those services can include prompts, schema metadata, query context, selected result rows, and workspace configuration necessary to complete the requested task.
- We do not train Embrasure-owned foundation models on customer data.
- We do not sell customer data or share it for cross-context behavioral advertising.
- For listed AI subprocessors, we configure commercial/API processing settings intended to prevent provider training on customer data and enable zero data retention where available.
- Customers can choose available model providers and should treat AI outputs as generated assistance that may require review before business use.
Cookies and analytics
We use necessary cookies and similar technologies for authentication, session management, security, preference storage, and load balancing.
We use our own analytics to understand how the website and product are used, diagnose reliability issues, and improve onboarding and core workflows. It stores a random identifier in your browser's local storage and a session identifier in session storage.
We use PostHog to record sessions on our homepage and in the signed-in product so we can find and fix usability problems. Recordings capture page layout and interactions. Form inputs are always masked. In the product, all text, images, and data values are also masked, page addresses are reduced to the product area, and recordings are linked to the signed-in user's ID, name, and email. Recording identifiers are kept in session storage and cleared when the tab closes. To opt out, block posthog.com in your browser or contact privacy@embrasure.ai.
We do not use cookies to sell personal information or to serve third-party behavioral advertising. Browser settings may allow you to block or delete cookies, but some product features may not work without necessary cookies.
How we disclose information
| Subprocessors | We use subprocessors to host, secure, deliver, support, and operate Embrasure. Our current subprocessors, including AI model providers, are listed on the Subprocessors page. |
|---|---|
| Customer admins | Workspace owners and administrators may access workspace membership, roles, audit events, settings, usage, content, and connected-source configuration according to their permissions. |
| Legal | We may disclose information when required by law, subpoena, court order, government request, or to protect rights, safety, security, and integrity. |
| Business transfers | If Embrasure is involved in a merger, acquisition, financing, reorganization, or sale of assets, information may be transferred as part of that transaction subject to appropriate protections. |
| With consent | We may disclose information when a customer or user directs us to do so or gives consent. |
We do not sell personal information. We do not share personal information for third-party cross-context behavioral advertising.
For information received from Google APIs, the Limited Use rules take precedence over the general disclosure terms above. Transfers are limited to consented, visible product features, security needs, legal obligations, or a merger, acquisition, or asset sale with the user's explicit prior consent. Human access is limited to specific data the user affirmatively permits us to view, security investigations, legal obligations, or aggregated data for lawful internal operations.
Security
We use administrative, technical, and organizational safeguards designed to protect personal information, including encryption in transit, encryption at rest, workspace access controls, row-level security, audit logging, least-privilege production access, and vulnerability management. More detail is available on our Security page.
No method of transmission or storage is perfectly secure. Customers are responsible for managing their own accounts, user permissions, connector scopes, and source-system access controls.
Retention and deletion
| Workspace data | Retained while the workspace subscription or evaluation remains active, then deleted or de-identified within 30 days after termination unless a contract, legal obligation, security need, or backup lifecycle requires a different period. |
|---|---|
| Backups | Backups are retained for disaster recovery and security purposes and expire on a rolling schedule. Deleted data may remain in backups until the backup expires. |
| Logs | Security, audit, and operational logs are retained as needed for investigation, abuse prevention, compliance, and reliability. |
| Commercial records | Invoices, contracts, tax records, and related business records are retained as required for accounting, legal, and compliance obligations. |
Customers may request deletion or export through their workspace administrator or by contacting us. We may need to verify authority before acting on a request.
International transfers
Embrasure operates primarily from the United States. If personal information is transferred from the European Economic Area, United Kingdom, Switzerland, or another jurisdiction with transfer restrictions, we use appropriate safeguards such as contractual commitments, data processing terms, and Standard Contractual Clauses where applicable.
Privacy rights
Depending on your location and relationship with Embrasure, you may have rights to access, correct, delete, restrict, object to, or receive a portable copy of personal information. You may also have the right to withdraw consent or lodge a complaint with a data protection authority.
If your information is part of a customer workspace, we may refer your request to the customer or ask the customer to authorize the request. To exercise rights, email privacy@embrasure.ai. We will respond as required by applicable law.
Legal bases for EEA, UK, and Swiss users
| Contract | To provide Embrasure, administer accounts, support users, and perform customer agreements. |
|---|---|
| Legitimate interests | To secure, improve, market, and operate Embrasure, prevent abuse, debug issues, and communicate with business contacts. |
| Legal obligation | To comply with accounting, tax, security, sanctions, law enforcement, and regulatory obligations. |
| Consent | Where required for optional communications, certain cookies, or other processing that depends on consent. |
California and other US state notices
Embrasure is a business-to-business service. We do not sell personal information and do not share personal information for third-party cross-context behavioral advertising. We use sensitive personal information only as needed to provide, secure, support, and improve the service, comply with law, or as otherwise permitted by applicable law.
Depending on your state, you may have rights to access, delete, correct, or receive a copy of personal information, and to appeal certain decisions. You can exercise those rights by emailing privacy@embrasure.ai.
Children
Embrasure is not directed to children and is intended for business use. We do not knowingly collect personal information from children under 16. If you believe a child provided personal information to Embrasure, contact us and we will take appropriate steps to delete it.
Changes to this policy
We may update this policy from time to time. If changes are material, we will provide notice through the product, by email, or by other reasonable means. The “last updated” date shows when this policy was most recently revised.
Contact
Privacy requests and questions: privacy@embrasure.ai
Legal notices: legal@embrasure.ai