Skip to app content
Skip to main content

Trust Center

How we protect your data and your ad accounts

Synter runs autonomous agents against real ad accounts and real budgets without hoarding your customer records. This hub links the Zero-Data-Retention controls, sub-processors, legal agreements, and live system status that back every workspace.

At a glance

The short version. Full technical detail is on our security & Zero-Data-Retention architecture page.

Zero-Data-Retention reads

Connected warehouse (BigQuery, Snowflake, Databricks, ADLS) and CRM (HubSpot, Salesforce, Attio) rows are read in volatile memory to extract context and never stored in Synter Postgres.

Encryption everywhere

OAuth tokens, API keys, and service credentials are encrypted at rest with AES-256. All traffic runs over TLS 1.3. We never store your ad platform passwords.

Workspace & sandbox isolation

Every customer's credentials, budgets, and agent threads run in an isolated workspace boundary, with dedicated Daytona cloud sandboxes for opt-in background bots.

Hard spend limits & audit receipts

New campaigns launch paused, server-side budget caps enforce hard ceilings, and every mutation writes an HMAC-SHA256 signed receipt with zero customer PII.

Internal security controls

Operational practices we run continuously, independent of any single feature or product release.

Access reviews

Repository and cloud infrastructure access is reviewed on a recurring basis, with least-privilege permission levels tracked per principal.

Branch protection & independent review

Staging and production branches require automated test verification and independent engineering review before promotion.

Vulnerability scanning

Dependencies and containers are scanned continuously with automated security audits and tracked remediation SLAs.

Secret management

Production credentials and encryption keys are governed through Google Secret Manager and Doppler with runtime injection into Google Cloud Run services and zero secrets in tracked files.

Environment separation

Development sandboxes (Daytona), staging, and production run in isolated environments with separate credentials, databases, and deployment pipelines.

HIPAA & Strict ZDR enforcement

Workspaces can enable Strict ZDR to scrub first-party pixel and form PII at rest, or HIPAA BAA mode (CUSTOM plan) to also strip stored URL query strings and suppress non-BAA subprocessors.

Sub-processors & Infrastructure

We use a focused set of vetted vendors to host, secure, and operate the service, including Google Cloud Platform (Cloud Run) and Railway for compute, Daytona for opt-in isolated bot sandboxes, Prisma Postgres for encrypted state, Clerk for authentication, and Stripe for payments. Every sub-processor and its HIPAA BAA mode posture is published on our sub-processors page.

View the full sub-processor list ->

System status

Live, real-time status for the API, database, MCP server, OAuth connections, and AI agent runtime.

View live system status ->

Data handling & policies

We do not sell personal data. Connected CRM and warehouse rows are never retained at rest, tactical agent memories are scrubbed of PII and ePHI before storage, and workspace data is deleted on account closure per our published retention schedule.

Security review, DPA, or HIPAA BAA request?

Our team responds to security questionnaires, executes DPAs and CUSTOM-plan HIPAA BAAs, and configures Strict Zero-Data-Retention or SSO for enterprise workspaces.

Contact Security Team
Trust Center | Synter