Subprocessors

Subprocessors and service providers.

This page identifies the core service providers Puddle may use to run authentication, hosting, live video, recording, transcription, speech, and AI-assisted review workflows.

Last updated: September 30, 2026

Current Service Providers

  • Amazon Web Services: cloud hosting, networking, storage, database, container registry, logs, secrets, and related infrastructure, including processing Calendar discovery samples and selected examples and corrections with Anthropic models through AWS Bedrock for interview discovery and user-reviewed recognition-rule suggestions. These features use US inference profiles and can process data across supported AWS regions within the United States.
  • WorkOS: customer authentication, session management, and account invitation workflows.
  • LiveKit: live audio/video rooms, participant connectivity, agent dispatch, and recording or egress workflows where enabled.
  • Anthropic: AI-assisted rubric review, follow-up question generation, ranking or recommendation support, and related model processing where enabled.
  • Deepgram: speech-to-text processing where enabled.
  • Cartesia: text-to-speech processing where enabled.

Data Handled

Subprocessors may process account identifiers, workspace data, candidate invite details, audio, video, transcripts, session events, device diagnostics, customer-provided rubrics, AI prompts and outputs, selected Calendar example titles, labels, durations, substituted organizer and recurring-series identifiers, corrections, logs, and support information depending on how the customer configures Puddle.

Updates

Puddle may update this list as service providers change. Customer agreements may include additional notice, objection, regional hosting, or data-transfer terms.

Questions about these terms or notices can be sent to hello@usepuddle.com. Candidates can also use that address to ask for accommodation, deletion, or an alternative process.