Skip to main content
Use this page to contact Resistant AI support for service issues, integration questions, and security incidents.

​
Contact channels

​
Before you contact us

  • Confirm the environment (testing vs production) and region/cell
  • Confirm you are using the correct API base URL and token URL
  • Confirm token issuance works (HTTP 200 + expires_in)
  • Confirm the upload succeeded (HTTP 200 from the presigned upload_url)
  • Confirm you’re polling the correct result endpoint(s) for your pipeline configuration
  • For UI/iFrame issues, confirm domain allowlist / CSP configuration
  • For events/webhooks, confirm your receiver/consumer is running and logging

​
What to include in your request

The fastest way to resolve issues is to provide a reproducible example with identifiers and timestamps.

​
Always include

  • Environment + region/cell (e.g., prod us-1)
  • Endpoint(s) and HTTP method (e.g., GET /v2/submission/{id}/fraud)
  • Timestamp(s) with timezone
  • submission_id (and query_id if used)
  • HTTP status code(s) and response body
  • Any request/correlation IDs from your logs (if available)

​
If you use SQS events

  • Queue ARN and region
  • Role ARN used for publishing
  • CloudWatch errors around AssumeRole / publish attempts

​
If you use webhooks

  • webhook-id, webhook-timestamp, webhook-signature headers
  • Receiver logs (status codes you returned, latency, retries)

​
If the issue is in the Web UI / Offline iFrame viewer

  • Web UI URL you accessed (or iframe host URL)
  • Browser + version
  • Screenshot of the error (ensure no PII is visible)
  • For iFrame: browser console output related to CSP or postMessage

​
Data and PII safety

Do not include PII in emails, ticket text, identifiers, screenshots, or logs. Common examples: emails, phone numbers, full names, addresses, or raw documents.
  • Do not email raw documents unless requested through an approved secure process.
  • If you need to share a sample file, use a non-production test document and confirm the sharing method with support.

​
Severity guidance

When reporting an issue, state the impact level:
  • Critical: production outage or system-wide failure
  • High: major degradation affecting many submissions or workflows
  • Normal: isolated integration issue or intermittent failures
  • Low: questions, configuration requests, feature requests

​
What happens next

  • Support will open a ticket and respond in the same thread.
  • We may ask for additional logs or a reproducible example.
  • If the issue is urgent, request escalation in the ticket or via your Customer Success contact.