feat: Connect Uploaded Documents (like CSVs and Excel files) to Code Interpreter for Data Analysis #4957

Closed
opened 2025-11-11 16:08:01 -06:00 by GiteaMirror · 1 comment
Owner

Originally created by @17jmumford on GitHub (Apr 24, 2025).

Check Existing Issues

  • I have searched the existing issues and discussions.

Problem Description

A common feature requested by many people is the ability for OpenWeb UI to do data analysis, similar to ChatGPT. Unfortunately, uploading a CSV and asking the LLM to help you analyze it leads to disappointing results.

Image

Desired Solution you'd like

I think we could combine two of the existing features to make this happen:

  • Document upload
  • Code interpreter
    If uploaded documents are made available in the code interpreter environment, then the LLM would very easily be able to analyze it!

More advanced capabilities are of course dependent on whether you are using a model with native tool use, but I think that we could make that happen.

Alternatives Considered

I saw that someone online had gone and created a whole alternative approach where they made a tool that is calling a docker file running a code interpreter.... seems like a very long-winded and brittle process. It would be nice to have this built right into OpenWeb UI instead.

Additional Context

It appears that there is some commented out code in the pyodide file related to exposing these documents

Image

@tjbck I see this commit from you.... do you want some help making this happen?
9c0a783991

Originally created by @17jmumford on GitHub (Apr 24, 2025). ### Check Existing Issues - [x] I have searched the existing issues and discussions. ### Problem Description A common feature requested by many people is the ability for OpenWeb UI to do data analysis, similar to ChatGPT. Unfortunately, uploading a CSV and asking the LLM to help you analyze it leads to disappointing results. ![Image](https://github.com/user-attachments/assets/b4ad00d3-bcdb-45a4-b755-48853a6df6e1) ### Desired Solution you'd like I think we could combine two of the existing features to make this happen: - Document upload - Code interpreter If uploaded documents are made available in the code interpreter environment, then the LLM would very easily be able to analyze it! More advanced capabilities are of course dependent on whether you are using a model with native tool use, but I think that we could make that happen. ### Alternatives Considered I saw that someone online had gone and created a whole alternative approach where they made a tool that is calling a docker file running a code interpreter.... seems like a very long-winded and brittle process. It would be nice to have this built right into OpenWeb UI instead. ### Additional Context It appears that there is some commented out code in the pyodide file related to exposing these documents ![Image](https://github.com/user-attachments/assets/70089e93-cadd-49c6-8e53-6266de987150) @tjbck I see this commit from you.... do you want some help making this happen? https://github.com/open-webui/open-webui/commit/9c0a783991077d460b526dc071841ae3b26c1f83
Author
Owner

@tjbck commented on GitHub (Apr 25, 2025):

Please check for existing issues and discussions first.

@tjbck commented on GitHub (Apr 25, 2025): Please check for existing issues and discussions first.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#4957