Subprocessors List

Last updated: 29 May 2026

This page lists the third-party service providers ("subprocessors") used by SC Technology (PTY) LTD to operate SC Scribe (the "Service"). Subprocessors may process personal information and Customer Content on our behalf solely to provide the Service.

For context: the Service is publicly accessible and does not require an account. Customer Content (audio, documents, transcripts, and generated minutes) is processed transiently by our application and is not stored in our application database or object storage after processing completes (request lifecycle). However, subprocessors may apply their own processing and retention practices under their own policies and settings.

Even with the controls described below, subprocessors may still process limited operational metadata and may retain data where required for security, abuse prevention, legal, or compliance purposes under their policies.

1) AI and transcription subprocessors

A) AssemblyAI

Purpose: Audio transcription (converting uploaded meeting recordings into text).

Data processed: Audio recordings uploaded by the user; transcription outputs (text).

Processing location: EU endpoint (configured by us via provider API base URL).

Notes: Audio is transmitted to AssemblyAI for transcription as part of providing the Service.

Retention controls: We use AssemblyAI's EU endpoint, our AssemblyAI account is opted out of model-improvement data sharing, and we issue transcript deletion requests once processing reaches a terminal state in our workflow.

B) Amazon Bedrock using Anthropic Claude Opus

Purpose: AI processing to generate meeting minutes, summaries, and action items from transcript text and any uploaded documents (agendas/prior minutes) used as context.

Data processed: Transcript text; uploaded documents provided for context; user instructions/prompts; generated minutes/outputs.

Processing location: AWS Europe (Ireland) source region using an EU geographic Amazon Bedrock inference profile. Under this configuration, prompts and outputs may be routed within EU AWS Regions for inference.

Notes: We use Anthropic Claude Opus through Amazon Bedrock for minutes generation. This regional processing posture supports our GDPR and POPIA alignment goals for users in South Africa, the United Kingdom, Europe, and elsewhere.

Retention controls: Amazon Bedrock is used for inference only as part of providing the Service. Providers may still process limited operational metadata and may retain data where required for security, abuse prevention, legal, or compliance purposes under their policies.

2) Hosting and infrastructure subprocessors

A) Amazon Web Services (AWS) - Africa (Cape Town) Region

Purpose: Hosting and infrastructure for the Service (compute and storage).

Data processed: Operational data required to run the Service; standard server logs; encrypted server volumes.

Processing location: South Africa (AWS Africa/Cape Town region).

Notes: We use encrypted storage on our server infrastructure.

B) Let's Encrypt

Purpose: TLS certificates to enable HTTPS connections to the Service.

Data processed: Domain validation information and certificate-related metadata (not meeting content).

Processing location: Varies by Let's Encrypt infrastructure.

3) Changes to subprocessors

We may update this list from time to time as the Service evolves. Where changes are material, we will update this page and the "Last updated" date.

4) Contact

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

support@sctechnology.co.za