This branch adds a demo program which makes use of the following:
Virtual Threads
Structured Concurrency
Pipelined Database Calls
Generative AI from Oracle Cloud
AI Vector Search with Oracle Database
The demo program will load a database table with text from a book full of facts about wild animals (the book is public domain). It will then query the text from the table, request embeddings for the text from OCI's Generative AI service, and store the embeddings back in the table as VECTOR data. Finally, it will perform a similarity search for some search terms using the VECTOR_DISTANCE function.
All steps of the program are executed using pipelined database calls on virtual threads. The virtual threads are managed using structured concurrency APIs which are being previewed in JDK 22.
Instructions for running the demo have been added to the README.
Michael-A-McMahon
commented
3 weeks ago
This branch adds a demo program which makes use of the following:
The demo program will load a database table with text from a book full of facts about wild animals (the book is public domain). It will then query the text from the table, request embeddings for the text from OCI's Generative AI service, and store the embeddings back in the table as
VECTORdata. Finally, it will perform a similarity search for some search terms using theVECTOR_DISTANCEfunction.All steps of the program are executed using pipelined database calls on virtual threads. The virtual threads are managed using structured concurrency APIs which are being previewed in JDK 22.
Instructions for running the demo have been added to the README.