Security & data

Know where your data goes. And how it is handled.

See how ProcessScout handles client evidence: where it is hosted, who can access it, how AI processes it and what happens when you need it removed.

Current service facts · Reviewed October 4, 2026

At a glance
Application hosting
DigitalOcean · New York, US
Web & API transport
HTTPS / TLS
AI privacy
Configurable masking of detected personal information
ProcessScout training datasets
Project consent required

Hosting & protection

Where the platform runs.

Application hosting and AI processing are separate parts of the service. A US-hosted application does not mean every connected provider processes data in the same region.

Application and file storage
Application infrastructure runs on DigitalOcean in New York. Uploaded files use DigitalOcean Spaces in its NYC3 region.
Encryption
Web and API connections use HTTPS/TLS. Stored AI-provider API keys are encrypted. File uploads are configured to request AES-256 server-side encryption; a requirement for encryption across every storage layer needs a specific deployment review.
Access to client work
Access is scoped through organizations, projects and roles. Client portal access requires an eligible organization and an active membership. Set the intended audience and inspect its view before sharing client material.

Understand the boundaries

AI and your data.

AI analysis uses project context to help investigate the work. Privacy controls, model selection and training consent determine how that context is handled.

01

Privacy before analysis

The default balanced setting masks detected personal information in AI-bound text, including names, contact details and payment identifiers. Projects can use different settings; full-detail mode disables this masking. Detection is not a guarantee that every sensitive detail is removed.

02

Training requires consent

Project data is excluded from ProcessScout training-dataset assembly unless an administrator grants training consent. AI interactions and redacted traces may still be retained to operate and review the service.

03

Know the model provider

Anthropic’s Claude is the primary AI provider for analysis and drafting. Sarah’s written answers use OpenAI, voice calls use xAI’s Grok, and search indexing runs on servers we operate. Provider selection can vary by feature or customer configuration.

What the primary provider says

Anthropic’s commercial terms say it does not train its models on inputs or outputs from its commercial API by default. Provider retention for abuse monitoring follows Anthropic’s published terms; zero-data-retention requires a confirmed provider arrangement and is not asserted for the standard ProcessScout service.

Read Anthropic’s commercial terms

Provider transparency

The services behind the experience.

These are the core providers described below. Additional connections and customer-configured AI providers can introduce other processors; confirm the full processing schedule for your engagement.

Core services and their purpose · October 4, 2026
ProviderPurposeData involved
DigitalOceanApplication hosting and file storageApplication records and uploaded evidence. New York infrastructure; Spaces NYC3.
Anthropic (Claude)Primary AI analysis, extraction and draftingContext sent for model processing under the selected privacy controls and provider terms.
OpenAISarah’s written answers and company researchContext sent for model processing under the selected privacy controls and provider terms.
xAI / GrokVoice conversations with SarahLive call audio and the context needed to answer it, under the selected privacy controls and provider terms.
ProcessScout-operated GPU serversSearch indexing (embeddings) and image readingText and images processed on infrastructure we operate; not sent to a third-party model provider for these steps.
StripeSubscription and assessment paymentsCheckout, billing and payment information.
RevvaloWebsite inquiries, chat and optional analyticsContact details and conversations; analytics follow your website privacy choice.

Connected business systems, email delivery and other configured services are additional parts of the processing review. This summary is not an exhaustive contractual subprocessor register.

Request the processing details for your scope

Retention & exit

Retention follows the type of data.

Project content, recordings, audit records and backups have different lifecycles. Agree the required retention and deletion scope for your workspace.

Project content

Retain evidence and working documents for the purpose and period agreed for your workspace. There is no universal 30- or 90-day deletion promise across the platform.

Recordings

Recording retention is configurable. A recording without a configured retention policy does not inherit an automatic deletion deadline; set the policy for the work you capture.

Audit records

The application’s default audit-log retention is 365 days and can be configured by organization. Audit records have a separate lifecycle from source evidence.

Deletion and account closure

Recording deletion tracks removal from storage. Organization closure begins with soft deletion and a seven-day grace workflow before permanent deletion. That grace period is not a promise that every copy is erased on day seven.

Derived outputs and backups

Source deletion, removal of derived material and backup expiry are distinct steps. Confirm the scope and completion schedule in writing when you need a specific deletion commitment.

A person to talk to

Bring the requirements. We’ll work through the details.

For a client security review, processing agreement or deletion request, contact [email protected]. Include your organization and the requirement; leave confidential evidence and credentials out of the initial message.

Build details
Build
os-bd3152dd322126ec
Built at
2026-10-16T00:50:39.151Z
Build record