Reusable Indexes

This April, we’ve introduced a major update to how data indexing works in deepset. Indexes are now standalone resources, fully independent from pipelines. Here’s what that means for you:

• Reuse data across pipelines. You only need to index your data once, then connect it to as many pipelines as you like.
• Simplify debugging. Indexing logs are now separate from pipeline logs, making it easier to track and troubleshoot.
• Share indexes across pipelines. Use the Document Store component to connect any index to your pipelines.

To help you get started, we’ve added a new Indexes section where you can manage all your indexes. You’ll also find ready-to-use templates to create indexes quickly and easily.

🎥 Watch the video to see the simplest way to create an index and connect it to a pipeline:

To understand indexes, check Indexes.

To learn how to create and enable indexes, see Working with Indexes.

Access to Multiple Organizations

Users can now be members of multiple deepset organizations. To add an organization and grant user access, reach out to your deepset representative.

When granted access, you can switch between your organizations using a list on the Organizations page:

For details, check Manage User Access.

Status Page

We have launched a deepset AI Platform Status Page to help you stay informed about the health and availability of our services.

You can now visit status.deepset.ai to check the general and search availability of the platform with breakdown per day. The page is connected to our monitoring system so the status is always up to date. Check also deepset AI Platform Status.