Provider abstraction
Choose the right provider for the job. A provider abstraction is intended to keep app experiences from being tied to one model vendor.
The best ecosystem feels connected without giving up control. JT AI Core is the planned shared intelligence layer for JT apps—designed to keep AI provider choices, model configuration, tool access, and privacy rules consistent while allowing each app to use only the capabilities it needs.
The best ecosystem feels connected without giving up control. JT AI Core is the planned shared intelligence layer for JT apps—designed to keep AI provider choices, model configuration, tool access, and privacy rules consistent while allowing each app to use only the capabilities it needs. The JT ecosystem is designed to connect products while keeping each workspace focused on its own job.
Choose the right provider for the job. A provider abstraction is intended to keep app experiences from being tied to one model vendor.
Keep configuration understandable. Model, endpoint, and connection settings should be explicit rather than hidden in application code.
Give each app only the tools it needs. App-scoped permissions are intended to limit unnecessary access and make consent easier to review.
Make privacy and diagnostics visible. Connection status, errors, and usage-related controls should help users understand what is happening.
Power a more connected JT. The interface is designed to make your next step easier to see, your decisions easier to revisit, and your work easier to organise.
Provider abstraction is intended to give JT apps a consistent way to request AI capabilities.
Model and endpoint settings should be configurable without exposing private keys in browser code.
Permission scopes and confirmation rules are intended to help keep tool access appropriate to each app.