Fragmented visibility
Understand which AI services are used, by whom, and what information is sent.
PROMPTLENS / SECURITY GATEWAY
Your organization’s security gateway to AI
Sensitive data detection, policy enforcement and visibility into AI traffic — on your own infrastructure.
Coverage applies to traffic routed through the gateway, on supported devices and services within the deployment.
Business partners
Space reserved for future partnersLOGOPartner name
LOGOPartner name
LOGOPartner name
LOGOPartner name
LOGOPartner name
LOGOPartner name
01 / CONTEXT
AI adoption grows. Oversight needs to keep pace.
Sensitive data in motion
Customer records, source code and organizational secrets may reach external services.
Evidence of control
Security and compliance teams need to explain policy decisions and reconstruct events.
02 / CAPABILITIES
From visibility to action. From action to evidence.
Usage visibility
Bring together prompts, responses and files from supported services, with organizational context.
Sensitive data detection
Identify personal and financial data, secrets, API keys, source code and custom keywords.
Policy enforcement
Log, alert, redact or block according to policies defined by your organization.
Investigation and access
Build cases, review evidence with permission and record a reason for viewing content.
Audit trail
Document actions, content access and policy acknowledgements for review and oversight.
Reporting
Usage, risk and departmental reports in PDF and Excel, in Hebrew and English.
03 / DATA PATH
Four steps. One control layer.
- 01
Route
Managed-device traffic is routed through the gateway using MDM configuration.
- 02
Detect
Content is checked for the sensitive data categories defined by your organization.
- 03
Enforce
The gateway logs, alerts, redacts or blocks according to policy.
- 04
Investigate
Review events in the console alongside access controls, audit records and reporting.
System components and licence verification operate locally. Using an external AI service still requires communication with that service; approved data may be sent according to policy.
04 / PRODUCT INTERFACE
One interface for day-to-day security work.
From organization-wide activity to a policy decision or content access - with clear context for each action.
Security dashboard
Review activity, events and findings by service and department.
Interface screenshot · illustrative dataData policies
Define conditions and actions for logging, alerting, redaction and blocking.
Interface screenshot · illustrative dataControlled content access
Viewing content requires permission and a reason, and is recorded in an access log.
Interface screenshot · illustrative data05 / PRIVACY BY DESIGN
Control stays on your infrastructure.
Self-hosted deployment separates risk detection from permission to view content. The product is designed for overt, authorized workplace monitoring.
Product privacy and access controls- Organization-owned devices and accounts, subject to policy and employee notice.
- No keystroke capture or screen recording, according to the product specification.
- Least-privilege access, a reason for viewing content and an access record.
- Encryption of content and files at rest and in transit.
For organizations that work with sensitive information.
Finance
Oversight of financial information and customer records.
Healthcare
Access controls and handling of sensitive personal information.
Legal
Control over documents, contracts and confidential information.
Industry and defense
Organizational data policies and separation of information.
06 / DEPLOYMENT
Start at the right scope. Expand with control.
Two months · up to 25 monitored users
Evaluate scenarios and policies in your environment.
Contact usSix months · up to 100 monitored users
Expand visibility and control across teams.
Contact usOne year · scope agreed separately
An organizational deployment aligned with your requirements.
Contact usQuestions before you get started.
Where is the system deployed?
PromptLens is designed for deployment on your own infrastructure. Configuration and coverage are defined during implementation.
How does traffic reach the gateway?
Managed devices route traffic through the gateway using organizational configuration distributed through MDM.
Which AI tools are covered?
The source material describes ChatGPT, Claude, Gemini and Copilot. Coverage depends on supported services and configuration; additional names in the diagram illustrate the architecture.
What happens when sensitive information is detected?
Depending on policy: logging, alerting, redaction of sensitive sections or blocking before transmission.
Who can view content?
Content access is subject to permission, a recorded reason and access logging.
Does everything work offline?
Licence verification is local. Requests to external AI services still require connectivity.
About the initiative
PromptLens is being built as an independent technology initiative to help organizations use artificial intelligence while maintaining control over the information sent to language models.
The system is intended to monitor traffic to language models, document usage and block transfers according to organizational policy. Our vision is to give organizations visibility and control over AI usage and help reduce exposure of sensitive information.
Product in testing
PromptLens is in the testing stage. Monitoring, usage records and blocking are being evaluated as part of testing; this does not indicate commercial availability or an open pilot.
Behind the product
The initiative was founded and is developed by Shmulik Tubul, an AI developer and engineer with extensive experience in information systems, IT infrastructure and technical support. The product combines practical experience in organizational environments with AI development expertise, grounded in an understanding of organizational needs and the challenges of adopting new technologies.
Shmulik Tubul
Founder & Product DeveloperFounder & AI Engineer
LET’S TALK
Let’s discuss AI usage in your organization.
Leave your details to arrange a meeting and explore PromptLens. Final wording and contact details will be added later.
Details are voluntary and used only to arrange a meeting. The form will open an email in your own mail application for you to review and send. The website does not store the details. Recipient and WhatsApp number will be added later.
Contact details will be available soon.