diff --git a/docs/src/pages/guide-document-qa.mdx b/docs/src/pages/guide-document-qa.mdx
index 97e58f94a3d72..2453f7d860e46 100644
--- a/docs/src/pages/guide-document-qa.mdx
+++ b/docs/src/pages/guide-document-qa.mdx
@@ -36,7 +36,7 @@ We will use a Data Source to automatically chunk, embed and index 4 IPCC reports
You can review this section without the need to actually create the data
source, and use the pre-filled public Data Source
- [ipcc-ar6](https://dust.tt/w/3e26b0e764/ds/ipcc-ar6) in the rest of the guide.
+ [ipcc-ar6](https://dust.tt/w/3e26b0e764/builder/data-sources/ipcc-ar6) in the rest of the guide.
You can create your first Data Source from your Dust home page by selecting the **Data Sources** tab
@@ -81,7 +81,7 @@ if it already existed for the provided identifier.
Once the 4 documents are inserted, you should be able to see them in the Data Source interface. You
-can review the public [ipcc-ar6](https://dust.tt/w/3e26b0e764/ds/ipcc-ar6) Data Source as an
+can review the public [ipcc-ar6](https://dust.tt/w/3e26b0e764/builder/data-sources/ipcc-ar6) Data Source as an
example.
Now that your Data Source is ready (or you plan to use
-[ipcc-ar6](https://dust.tt/w/3e26b0e764/ds/ipcc-ar6)), you can move on to the next section to create
+[ipcc-ar6](https://dust.tt/w/3e26b0e764/builder/data-sources/ipcc-ar6)), you can move on to the next section to create
a Q&A app that will use it.
## Creating a Q&A app
In this section we'll provide step-by-step instructions to create a Q&A app that will use the data
source to answer user-provided questions. You can find a finalized version of the app
-[here](https://dust.tt/w/3e26b0e764/ds/ipcc-ar6).
+[here](https://dust.tt/w/3e26b0e764/builder/data-sources/ipcc-ar6).
### `input` block
@@ -123,7 +123,7 @@ Data Source:
Once the `input` block points to your new dataset you can add a `data_source` block and select your
Data Source. If you want to use the publicly available Data Source
-[ipcc-ar6](https://dust.tt/w/3e26b0e764/ds/ipcc-ar6), edit the user name to `spolu` and then select
+[ipcc-ar6](https://dust.tt/w/3e26b0e764/builder/data-sources/ipcc-ar6), edit the user name to `spolu` and then select
`ipcc-ar6`. You should set `top_k` to 12, meaning that we'll consume 3072 tokens of context to
present chunks to the model.
@@ -146,7 +146,7 @@ At this point you can run your app and introspect the outputs of the `data_sourc
We will use `gpt-3.5-turbo` to process the chunks and generate an answer to the user.
We use the following instruction in the context of
-[ipcc-ar6](https://dust.tt/w/3e26b0e764/ds/ipcc-ar6):
+[ipcc-ar6](https://dust.tt/w/3e26b0e764/builder/data-sources/ipcc-ar6):
```txt
You are an helpful assistant. For each user question, you'll receive 12 additional system
diff --git a/front/components/data_source/DataSourcePicker.tsx b/front/components/data_source/DataSourcePicker.tsx
index 769c343621a9b..27e70280fd725 100644
--- a/front/components/data_source/DataSourcePicker.tsx
+++ b/front/components/data_source/DataSourcePicker.tsx
@@ -70,7 +70,7 @@ export default function DataSourcePicker({