-
Notifications
You must be signed in to change notification settings - Fork 117
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Upsert documents & tables on upload for JIT. #8698
Upsert documents & tables on upload for JIT. #8698
Conversation
2bcec63
to
444dae2
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks great!
709e334
to
c55c108
Compare
|
238db9e
to
330d600
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
A few comments from skimming through it
fc56fc2
to
dfafb75
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm not a big fan of introducing the JIT brick in the file API this way. IMHO, it seems more like a post-processing task rather than a pre-processing one. Is there a chance we could move this to a second step after the file has been uploaded?
6d5a92e
to
60e901f
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks very reasonable.
We don't handle yet the file upload and then conversation creation yet do we?
a1f5e28
to
b8964a7
Compare
🙏🏼
We do now (check last commit) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks way better this way, left a few comments.
b8964a7
to
35f5924
Compare
Description
To support JIT actions, we need to upsert the documents and / or tables to a dataSource in the conversations space.
File model changes:
useCaseMetadata
, to hold the conversation sid from which the file was uploadedid instead but that would mean fetching the conversation in all callers of the
processAndStoreFile
function based on useCase & metadata, which I feel are not their concerns.snippet
, to hold the LLM summarized description of the content (will fill in an upcoming PR).Client-side changes:
useCaseMetada
as an optional argument ofuseFileUploaderService
for both front and extensionuseCaseMetadata
to the dust-client typesBackend-side changes:
Next PRs:
Risk
Since everything is optional and behind a feature flag, it should not break anything.
Deploy Plan
Apply migration on prodbox, then deploy
front