Put policy at the decision point
Evaluate how defined sensitive patterns should be handled before an approved client submits text to an AI destination.
SentraMindby wGrow
A clear decision.
Inside a boundary you define.
SentraMind helps organisations evaluate how sensitive prompts should be inspected, how customer-defined policies should apply and how each outcome can be reviewed.
Platform Core + Sentry · Available for POC
A business concern behind the prompt
Customer information, credentials and internal material can enter everyday AI conversations. SentraMind provides a focused way to evaluate the controls and decisions around that exposure.
Evaluate how defined sensitive patterns should be handled before an approved client submits text to an AI destination.
Connect a warning or block to the customer-defined rule and policy version behind it, with a record reviewers can inspect.
Examine the processing boundary, information access, evidence handling and operating responsibilities as part of the POC.
A product with a defined starting point
The product separates today’s POC offer from planned capabilities. Review the current boundary before choosing the scope of an evaluation.
Available for POC
Customer-defined policies, controlled decision paths and reviewable records form the foundation of the current POC.
Policy and evidence foundationAvailable for POC
GenAI Prompt DLP: deterministic inspection of agreed text examples and customer-defined warning or block outcomes.
Prompt inspection and policy outcomesComing next
Planned AI usage observability to help organisations understand where AI is used and where controls may be needed.
AI usage visibilityDesign-partner roadmap
The intended private reasoning layer for specialist crews. The current build has no configured model runtime.
Private model and knowledge directionThe current POC boundary: endpoint capture is simulated and the model classifier is disabled. Production identity, integrations, throughput, latency and availability require further validation. Self-hosted and air-gap-capable operation are design directions whose complete deployment behaviour must be demonstrated.
Read the current platform scopeFrom an input to an accountable outcome
Deterministic checks and customer-defined policy govern the result. The control path makes the decision and the reason behind it inspectable.
An approved example enters the agreed inspection path.
Deterministic checks identify policy-relevant content.
The customer-defined policy determines the outcome.
Inspect the outcome and its policy reference.
Conceptual workflow. The current capture path is simulated; public browser demonstrations use synthetic examples to exercise deterministic behaviour.
Explore the demonstrationsExplore the product
SentraMind is built by wGrow. Its own website brings together the platform, demonstrations, deployment considerations and product roadmap.
Our engineering services address the wider business environment: infrastructure, security, AI adoption and the responsibilities around operation. Each product evaluation or service engagement has its own agreed scope.

Clarity before a commitment
A useful POC starts with the workflow, policy owner, information boundary and approved test examples. The evaluation then establishes expected outcomes, known exclusions and the evidence needed for the next decision.
Security and platform owners, governance teams and business leaders each have a role in agreeing that boundary. Integration and deployment readiness are assessed in the proposed environment.
Current intake status: private-evaluation intake is not yet enabled on the product website. The public demonstrations and preparation guide are available to explore now.
Know what you are exploring
SentraMind is a wGrow product for organisations evaluating AI-security controls in customer-controlled environments. Its current Platform Core and Sentry POC focus on deterministic prompt inspection, customer-defined policies and reviewable decisions.
Platform Core and Sentry GenAI Prompt DLP are available for a scoped proof of concept. The evaluation defines approved examples, policy outcomes, limitations and acceptance criteria. Current endpoint capture is simulated; browser and identity integrations require explicit scope and validation.
The public browser demonstrations exercise deterministic behaviours using synthetic examples. They help explain the control path and policy outcomes. Production capture, integrations, deployment capacity and operating controls must be established in the agreed environment.
Private AI Core is on the design-partner roadmap. The current build has no configured model runtime and the model classifier is disabled. Its published architecture and sizing guidance describe the intended direction.
SentraMind is the product identity, built by wGrow. wGrow’s infrastructure, security and AI FDE services address the wider engineering and operating work around a business. Any product deployment or service engagement needs its own defined scope and responsibilities.
Visit the SentraMind website, try the registration-free browser demonstrations and review the platform and deployment guides. Private-evaluation intake is not yet enabled on the product website; the evaluation guide explains the intended process and prerequisites.
SentraMind by wGrow
Explore the current product boundary and try the public browser demonstrations.