Security and data handling

Security begins with a documented operating model, not a collection of badges.

AlchemyWorkflow is delivered through managed enterprise access and a scoped implementation. The security design connects user roles, integration permissions, workflow controls, human approvals and execution records.

Important:

This page describes the intended control model and areas reviewed during implementation. It does not claim certifications, hosting locations, encryption specifications, recovery objectives or identity features that have not been confirmed for the production service.

Control principles

What is designed into each implementation.

01

Customer-managed access

Employees do not create public accounts. Access begins with customer administrator invitations and the roles configured for the tenant.

02

Scoped implementation

Data sources, allowed actions, approval boundaries and operational owners are defined during discovery and recorded in the implementation scope.

03

Minimum necessary data

Workflows should access and move only the records and fields required for the agreed process.

04

Human control for higher-impact actions

Customer-facing, financial, legal or otherwise sensitive actions can remain subject to review and approval.

05

Traceable execution

Workflow runs can preserve triggering context, decisions, connected actions, owner assignments, exceptions and completion status.

06

Shared responsibility

AlchemyWorkflow configures and operates the service within the agreed scope; customers remain responsible for their users, source data, connected-system authority and business decisions.

Implementation lifecycle

Security decisions are made before a workflow goes live.

1. Discovery

Identify the systems involved, data categories, process owners, risk points and required approvals.

2. Access design

Define administrators, workflow owners, operators, reviewers and users who only need visibility.

3. Integration review

Confirm authentication method, permission scope, fields used, actions permitted and failure behavior.

4. Workflow testing

Validate normal paths, duplicates, missing data, permission errors, unavailable systems and manual escalation.

5. Production release

Publish the approved configuration and provide customer administrators with the agreed operating controls.

6. Ongoing review

Review access changes, workflow revisions, exceptions and integration behavior as the operating model evolves.

Responsibility matrix

Controls need a named owner.

AreaExpected controlPrimary ownership
Identity and access

Administrator invitation; role allocation; removal of access when users leave or change responsibility.

Customer administrator with implementation support.

Workflow changes

Named ownership; controlled editing and publishing; testing before production release.

Authorized workflow owner and customer administrator.

Connected systems

Tenant-specific authorization; minimum required permissions; documented read and write actions.

Customer system owner and implementation team.

Approvals

Named reviewer or group; decision status; escalation when approval is delayed or rejected.

Customer-designated business owner.

Execution records

Trigger context; path selected; action status; exception and assigned owner.

Operational owners according to role.

Data lifecycle

Retention, export and deletion requirements documented in the applicable agreement and tenant design.

Customer and AlchemyWorkflow under the contract.

Data boundary

Define what the workflow needs—and what it should never receive.

During discovery, teams should document:

  • Data fields required for decisions and actions
  • Systems that remain the authoritative source of record
  • Categories of sensitive data excluded from the workflow
  • Approved recipients and downstream systems
  • Retention, export and deletion requirements
  • AI services, if any, permitted to process workflow context

Customers should not submit regulated or sensitive categories unless the applicable agreement, architecture and controls expressly support that data.

Security review FAQ

Claims we will not leave ambiguous.

Is AlchemyWorkflow SOC 2 or ISO 27001 certified?

This website does not claim either certification. Current assurance materials and any customer-specific security requirements should be discussed during the sales and security review process.

Does the platform support SSO or MFA?

Authentication requirements should be confirmed during discovery. They are not represented here as generally available features until the deployment and identity design are agreed.

Is each tenant a physically separate deployment?

The exact isolation model must be documented in the order, SOW or security documentation for the customer. The term dedicated tenant does not, by itself, promise a separate cloud account, database or infrastructure stack.

Where is customer data hosted?

Hosting region, infrastructure provider and any data residency requirement must be confirmed for the actual production deployment and reflected in the agreement and DPA.

Is customer data used to train AI models?

No blanket claim is made on this page. Any AI provider, input handling, retention and model-improvement terms must be documented for the services actually enabled in the customer tenant.

How are security incidents handled?

Incident contacts, notification commitments and responsibilities should be stated in the signed agreement and the operational procedures applicable to the production service.

Bring your security and data requirements into discovery.

We will identify the controls that must be confirmed before the implementation scope is finalized.

Contact Sales