sdsd sdsds
Agent - Fix messageplain
Add history to messages parse_sentance.json
Test Multi Step Search Find All the product... Now find recent products
Finish RAG
Semantic search
Show similar vector results. Search Products...?
UN bugs
AI Failures: Remote Steps did not come up in Vector Search all children when picking one type Plural vs Singular Collections
Bugs:
Dont repeat Semantic results Not filtering by collection Ajax order of messages from AI
Consolidate JPG usage Replication servers Galleries not keeping the check box Save All needs to make module api
Fix tool calling: https://www.reddit.com/r/LocalLLaMA/comments/1mje5o0/psa_qwen3coder30ba3b_tool_calling_fixed_by/
Some docs - Microsoft Auth works.
Cristobal, please setup a meeting with ALM and Ian to enable this
Ian, Yessenia with ALM wants to move the meeting for the Microsoft Auth config 30 minutes earlier, you are ok with that?
Should we prepare something in their instance?
Should just need to upgrade it so the correct Oauth code is there
Ian reported that switching to spanish did not work while giving a demo.
Was a 404 from the picker - see attached -
- Drop Documents in Submodule Collections > Documents not work
Issues: Delete pages when we delete a document microsoft OAUTH login is mixed in with google login Drag and drop duplication Drag and drop submodule files Add comma and link to Page 20Page34 Dont show others sessions? Make session public/shared checkbox with lock icon
Lost the plot: https://fay-informatics.emedialibrary.com/finder/find/views/modules/entitydocument/index.html?entityid=AZqUYSqdY55-Ccvh-Iva Channel in the session?
Shakil, I've decided to do my testing against this - it's very dense and complicated so I think maybe better than those ontario docs. How are you progressings?
We want legal docs over here https://fay-informatics.emedialibrary.com/finder/find/index.html as a testing platform
Shakil is out till next week. I can help you test if needed
That's OK - we were goign to try to compare llama-index vs my custom approaches. I just wanted to have a consistent document between him and I. i would like to know what backends he's using for the vector searches he's doing also - see how much faster or slower they are compared to the ones I wrote. I will keep you posted on my RAG process!
CloudTranscodeManager was deleted, - I know we don't do the whole depracate thing - but how do I queue an autoTranslate now?
