Sub-processors

Last updated: 1 September 2026

To deliver the Pensieve service we share data with the vetted third-party sub-processors listed below. Each is engaged under terms that require appropriate security and confidentiality, and that prohibit using your content to train their own models.

Pensieve sub-processors
Sub-processorPurposeLocation
SupabaseDatabase, authentication and file storageEuropean Union (AWS, Paris)
Google Cloud PlatformDatabase hosting for your context layerUnited Kingdom (London)
RailwayApplication hostingUnited States
NangoManaged OAuth and connector synchronisationUnited States
OpenRouterAI gateway — every model, embedding and web-search request is routed through itUnited States
OpenAIAI model processing and text embeddings, reached via OpenRouterUnited States
AnthropicAI model processing, reached via OpenRouterUnited States
FirecrawlWeb page retrieval for the website connector and the agent's web fetchesUnited States
StripeSubscription billing and payment processingUnited Kingdom and United States
MuxVideo hosting for demo and marketing pagesUnited States
PostHogProduct analytics and error trackingEuropean Union (Frankfurt)
OnePatchApplication performance monitoring — traces, metrics, logs and error diagnosticsUnited States
ResendTransactional emailUnited States

International transfers

Our primary data stores — the database, file storage and your context's context layer — are hosted in the UK/EU. Some sub-processors (in particular our AI model providers) are based in the United States, so delivering the service involves transferring data outside the UK/EU. Where this happens we rely on Standard Contractual Clauses and equivalent safeguards.

Changes

We will update this page before engaging a new sub-processor and aim to give at least 30 days' notice of material changes. Questions? Contact us at euan@pensieve.uk.

See also our Privacy Policy.