The platform that runs your supply chain decisions meets the standards your security team expects, with the certifications, controls, and documentation already in place.
Security reviews shouldn't stall supply chain AI. Every control is enforced at the platform level, so every app, assistant, and agent inherits the same posture automatically.
Every customer environment is isolated. Your data is never co-mingled with anyone else's, and never used to train shared models.
Single sign-on, role-based access, and row-level security, with every action recorded in logs your security tools can read.
Every answer and recommendation traces back to its source data. Autonomy is delegated by you, never assumed, and can be paused at any time.
Independent audits, regular penetration testing, and continuous vulnerability scanning, with documented controls behind each one.
Choose where ConverSight runs. Either way, security is built into the platform, and the data is yours.
OPTION 01
A private, isolated environment in our secure cloud, with regional data residency options.
OPTION 02
Deploy inside your own hosted environment, so data never leaves infrastructure you control.
EITHER WAY
Add your own sources, build your own APIs, and take everything with you if you ever move on.
The same controls run through all four layers, from the data to the decision. Pick a layer to see what it enforces.
LAYER 1
Data and knowledge
Encrypted at rest (AES-256) and in transit (TLS 1.3)
Full tenant isolation
Role, user, and row-level permissions, set once
Regional data residency and customer-managed keys
LAYER 2
AI assistant and analyst
Sees only what the person asking can see
Claude and ChatGPT connect through a governed MCP layer
Bring your own LLM, or use the closed-loop native model
Your data never trains shared models
LAYER 3
Apps and blueprints
Build rights granted by role
Changes tested in a sandbox before production
Draft, approve, and publish, with every version kept
Apps inherit Digital Cube permissions
LAYER 4
Agents and approvals
Approval rules by value, supplier, region, or site
Routine actions close only under policies you set
Every decision in the Audit Trail, with its reasoning
Pause or stop any agent, at any time
When someone asks a question from Claude or ChatGPT, the request goes through ConverSight's governed MCP layer. The assistant gets the answer that person is allowed to see, nothing more. It never gets direct access to your data lake.
Each person's permissions apply, on every request
A proprietary security layer blocks anything outside them
Prefer to keep everything in-house? Use the closed-loop native model
A person asks a question
01
Checks who's asking and what they may see
02
Governed data and business logic
03
↺ Only the permitted answer goes back.
⛨ Security controls
What your security team will ask about, in one place.
Private cloud deployment
Network segmentation
Zero-trust architecture
24/7 monitoring
AES-256 encryption at rest
TLS 1.3 encryption in transit
Tenant isolation
Customer-managed keys
Single sign-on
Role-based access
Row-level security, including for suppliers
Activity logs for your security tools
Backups across isolated locations
Redundant systems with automatic failover
Capacity planning
Tested recovery procedures
Regular penetration testing
Continuous vulnerability scanning
Scheduled, tested patching
Ongoing risk assessments
Background checks
Security training
Least-privilege, need-to-know access
We never sell or mine customer data
Everything your review needs. Reports are shared under NDA.
Independent audit of our controls
Information security management
Send yours, or start from our completed one
GDPR terms for customer data
Still have a question? Our security team answers questionnaires directly.
ConverSight is SOC 2 Type II and ISO 27001 certified, and GDPR compliant. Audit reports are available under NDA from the security documents section above.
Either in ConverSight’s secure cloud, in an isolated environment with regional residency options, or in your own cloud. Security is built into the platform either way.
You do. You can add your own sources and build your own APIs on the data lake, and if you move on, the data goes with you.
No. Your data is never used to train shared models, and it’s never co-mingled with other customers’ data.
Requests go through ConverSight’s governed MCP layer, which applies the asking person’s permissions and returns only what they’re allowed to see. The outside assistant never gets direct access to your data lake. If you’d rather keep everything in-house, use ConverSight’s closed-loop native model instead.
Yes. Roles can restrict data by region, product, customer group, or any other dimension, for internal teams and for external users such as suppliers. The same rules apply in every app, in Athena IQ, and in every agent.
Yes. Agents only act within the approval rules you set, any agent can be paused at any time, and every action is recorded with who approved it, or which policy allowed it.
Yes. Activity logs can be sent to your existing security monitoring tools, so ConverSight shows up alongside everything else your team watches.
