Contracts hold pricing, liabilities, personal data and commercial strategy. Many organisations are understandably wary of putting all of that into someone else’s cloud. So we made a deliberate choice: ContractSense runs in your environment.
What runs where
The full platform, including the web app, the lifecycle API, the intelligence service and the background workers, is installed in your environment. You can run it on a single server or in your own Kubernetes cluster. All contract records, versions, files, search indexes and the audit trail stay there.
We host one thing: the packs server. It serves versioned industry content (clauses, templates, prompts and skills) and keeps a ledger of AI usage and cost.
How the connector works
Sometimes a hosted service needs to read a contract, for example to apply an industry pack. It never connects into your network. Instead:
- It asks for a contract read through the packs gateway.
- The gateway checks the token and scopes, logs the usage and queues a job for your tenant.
- Your connector polls outward over HTTPS, claims the job and runs a fixed operation with that user’s permissions.
- The result returns by the same route.
If no connector has polled for two minutes, the call fails immediately. You open no inbound port, and the connector can only run fixed, pre-defined operations.
Content over code
Because industry knowledge arrives as packs, adding a sector does not mean changing your installation’s code. You get new clauses, templates and prompts as content, with versions, while your data stays put.
Being straight about the roadmap
We would rather be clear than oversell. Single sign-on, multi-factor authentication and platform-level encryption at rest are on our roadmap. Ask us for current status. Read more on our security page.