How do we know which Knowledge Bases provider we a...
# support
s
I saw the notification on our dashbaord
Copy code
We are permanently shutting down support for Canonical and migrating Knowledge Bases to Trieve!
How do we know which one we are currently using? And if we are using the provider from Canonical, how do we migrate it to Trieve?
a
Heysmoothie Thank you for creating the ticket! A member of our team will reach out to you shortly.
v
@smoothie I'm glad you reached out for clarification on the migration process from Canonical to Trieve. To determine which Knowledge Base provider you are currently using, you would typically check your service configuration or the administrative settings of your dashboard. This information is usually found in the section where integrations or third-party services are listed. If you find that you are using Canonical and need to migrate to Trieve, the general steps for migration would include: 1. Data Export: Export your data from the Canonical Knowledge Base. This often involves using an export tool or feature provided by Canonical. 2. Data Transformation: Ensure that the exported data is in a format compatible with Trieve. You may need to convert the data or adjust it to meet Trieve's import requirements. 3. Trieve Account Setup: Set up your account with Trieve if you haven't done so already. This will include configuring the basic settings and preferences for your new Knowledge Base. 4. Data Import: Import your transformed data into Trieve. Trieve will likely have documentation or tools to facilitate the import process. 5. Verification and Testing: After importing, verify that all data has been transferred correctly and test the Knowledge Base to ensure it's functioning as expected. 6. Update Integrations: Update any integrations or API endpoints in your applications to point to Trieve instead of Canonical. 7. Monitor and Support: Monitor the new setup for any issues and provide support to users during the transition. It's important to consult the official migration guide provided by Trieve or reach out to their support team for specific instructions and assistance. They may have tools and services designed to streamline the migration process. If you need further assistance with the migration or have any other questions, please let me know!
You can continue this conversation by mentioning me in the message.
s
@Vapi if we are using Vapi's UI to upload the files, how do I know which provider we were using?
v
@smoothie If you are using VAPI's user interface to upload files, the Knowledge Base provider you are using might not be directly visible in the UI. To determine which provider you were using, you would typically look for any provider-specific settings, documentation, or references within the VAPI dashboard or configuration pages. Here are a few steps you can take to identify your current Knowledge Base provider: 1. Check the Settings or Configuration Area: Look for any sections in the VAPI dashboard that pertain to Knowledge Base settings or integrations. The provider's name might be listed there. 2. Review Documentation: If VAPI has provided documentation or help articles related to Knowledge Base integration, it may mention the supported providers. 3. Examine API Endpoints: If you have access to the API configuration, the endpoints might indicate which provider is being used. For example, the URL structure might include the provider's name. 4. Contact Support: If you're still unsure, the most reliable way to find out is to contact VAPI support directly. They can provide you with information about your current setup and guide you through the migration process if necessary. Would you like me to assist you with contacting support to clarify which Knowledge Base provider you are currently using?
Got it. Tagging @User for additional help.
j
We are permanently shutting down support for Canonical and migrating Knowledge Bases to Trieve! ..... Based on this I believe we are using Trieve for our Vector Database.
s
we starting using VAPI last week, so we don't need to do anything is that correct?
c
Lol VAPI uses trieve natively you dont need to do anything. This is old. Announcement is new.
s
Got it, thank you 😆
l
When we create assistant via API, we have to add knowledgeBase to the payload, and in that payload we currently use "knowledgeBase": { "topK": 2, "fileIds": [ "ID" ], "provider": "canonical" } VAPI has provided no information on what that payload should look like now, and if you use trieve for the "provider" it throws an error, does any one have any clarity on this?
c
Not sure how you can type this without reading the docs which have been out for weeks https://docs.vapi.ai/knowledgebase
l
Hey Chowderr, be nicer
You can create assistants via API and the docs don't explain how to use knowledge base, and historically we have just inserted file IDs: https://docs.vapi.ai/api-reference/assistants/create
c
I am super nice lol. if you want to attack my character please do better. There have literally been docs for weeks. You may just be seeing it - this change has been here for almost 3 months now. VAPI just announced it. Besides that saying there is nothing when links are right there is just diabolical work.
s
@smoothie checking if this issue is resolved for you.
@smoothie because of no response closing this ticket.
i
k
Hey, Can you use the automation that I wrote? It is pretty simple to use: 
c
Been in here not even 24 hours. It can be found with a simple search: https://docs.vapi.ai/knowledgebase. Be for real.
3 Views