Back to Trust Center
Contents

Subprocessor List

Effective Date: 1 April 2026 Last Updated: 1 April 2026

GTMSAI Ltd (trading as Pebbles Ai) uses the following third-party subprocessors to assist in providing the Service. This page provides transparency about who processes data on our behalf, what data they process, and where that processing takes place.

This list is maintained in accordance with our obligations under the UK GDPR, EU GDPR, and our Data Processing Agreement.


Current Subprocessors

Subprocessor Purpose Data Processed Location Website
Google Cloud Platform Infrastructure hosting (compute, storage, databases) All user data (application data, account data, user content) Belgium (europe-west1) cloud.google.com
Stripe Payment processing Payment details, billing information, transaction records US / EU stripe.com
HubSpot Sales CRM, marketing automation Contact information, email addresses, usage data, marketing interactions US (EU data centre available) hubspot.com
SendGrid Email delivery Email addresses, email content, delivery metadata US sendgrid.com
Google Analytics Website analytics Usage data, IP addresses, device information, page interactions US analytics.google.com
Microsoft Clarity User behaviour analytics (heatmaps, session replays) Session data, interaction data, device information US clarity.microsoft.com
Cloudflare CDN, DDoS protection, bot management Request data, IP addresses, traffic metadata Global (edge network) cloudflare.com
Google Workspace Internal business operations Internal communications EU workspace.google.com
GitHub Source code hosting, version control No user personal data US github.com
Sentry Error tracking and performance monitoring Error reports (may include user identifiers, IP addresses, request metadata) US sentry.io
Pusher Real-time message delivery (WebSocket) for streaming AI-generated content to users AI-generated content (which may reference contact data and email content in context), user/tenant identifiers in channel names, event metadata EU (Ireland, eu-west-1 cluster) pusher.com
Weaviate Vector database for AI-powered search and retrieval Vectorised embeddings derived from user content (contacts, organisations, conversation context) Germany (europe-west3, Frankfurt), Weaviate Cloud weaviate.io

AI/LLM Service Providers

Our AI features utilise multiple third-party large language model providers via API. The specific provider used may vary by feature and task. These providers process data transiently to generate responses and are contractually prohibited from using customer data for model training. Data is transmitted via encrypted connections.

Subprocessor Purpose Data Processed Location Website
Anthropic AI language model (Claude) for copilot and content generation User prompts, campaign content, contact context (as submitted by user) US anthropic.com
OpenAI AI language model (GPT) for copilot and content generation User prompts, campaign content, contact context (as submitted by user) US openai.com
Google Cloud Vertex AI (Gemini) AI language model (Gemini) for copilot and content generation via Vertex AI User prompts, campaign content, contact context (as submitted by user) US (us-central1 default; migration to europe-west1 planned) cloud.google.com/vertex-ai
xAI AI language model (Grok) for copilot and content generation User prompts, campaign content, contact context (as submitted by user) US x.ai
Perplexity AI AI-powered research and information retrieval Search queries, research context US perplexity.ai

These providers do not persistently store personal data beyond the duration required for abuse monitoring and safety purposes, as governed by our data processing agreements with them.


Data Protection Safeguards

All subprocessors are bound by data processing agreements that require them to:

  • Process personal data only on our documented instructions
  • Implement appropriate technical and organisational security measures
  • Assist us in fulfilling our obligations to data subjects
  • Delete or return personal data upon termination of the agreement
  • Submit to audits and inspections as required
  • Not use customer data for model training (AI/LLM providers specifically)

Where a subprocessor is located outside the European Economic Area or United Kingdom, we ensure that appropriate safeguards are in place, including Standard Contractual Clauses (SCCs) approved by the European Commission and/or the UK International Data Transfer Agreement (IDTA) addendum. For further details on international data transfers, see Section 11 of our Privacy Policy.

DPA and Transfer Mechanism Status

Subprocessor DPA EU SCCs UK Addendum No-Training Clause
Google Cloud Platform (incl. Vertex AI) Google Cloud Data Processing Addendum (CDPA) Included in CDPA Included in CDPA Yes (Section 17 Service Specific Terms)
Anthropic Commercial API Terms + DPA Available via Anthropic Trust Center Available via Anthropic Trust Center Yes (commercial terms)
OpenAI API Terms of Service + DPA Available via account settings Available via account settings Yes (since March 2023, API data not used for training)
xAI Enterprise DPA (x.ai/legal/data-processing-addendum) Module 2 (C2P) + Module 3 (P2P), Irish law UK Addendum included Yes (processor-only, no sell/share)
Perplexity AI DPA (perplexity.ai/hub/legal/dpa) Module 2 (C2P) + Module 3 (P2P), Irish law UK Addendum included Yes (explicit: not used for LLM training)
Pusher (Bird) DPA (bird.com/legal/dpa) Module 2 (C2P) + Module 3 (P2P), Dutch law UK Standard Contractual Clauses Yes (processor-only, no sell/share)
Stripe Stripe DPA Included Included N/A
SendGrid Twilio DPA Included Included N/A
HubSpot HubSpot DPA Included Included N/A
Cloudflare Cloudflare DPA Included Included N/A
Sentry Sentry DPA Included Included N/A
Weaviate Weaviate Cloud Terms of Service EU-hosted (Frankfurt) — no transfer outside EEA N/A (EU-only) N/A (vector database, not AI training)

Last verified: April 2026


Notification of Changes

GTMSAI Ltd will notify customers of any changes to this subprocessor list in accordance with the following process:

  1. Advance notice. We will provide at least 30 days' written notice before engaging a new subprocessor or materially changing the role of an existing subprocessor. Notice will be provided via email to the account owner and/or through an update to this page.

  2. Details provided. The notification will include the name of the subprocessor, the purpose of processing, the categories of data processed, and the location of processing.

  3. Right to object. If you have a reasonable objection to a new or changed subprocessor based on data protection grounds, you may raise your objection in writing to data@trypebbles.ai within the 30-day notice period. We will work with you in good faith to address your concerns, which may include offering an alternative configuration or workaround. If we cannot reasonably resolve your objection, you may terminate the affected services by providing written notice, and we will refund any prepaid fees for the unused portion of the subscription term following the effective date of termination.

  4. Deemed acceptance. If you do not object within the 30-day notice period, you are deemed to have accepted the new or changed subprocessor.


Staying Informed

To receive notifications about subprocessor changes, ensure that the email address associated with your account is current. You can update your email address in your account settings at app.trypebbles.ai.


Contact

If you have questions about our subprocessors or data processing practices, please contact:

Data Protection Officer: data@trypebbles.ai

Postal address: GTMSAI Ltd 82 Wandsworth Bridge Road London, SW6 2TF United Kingdom