- Move embedding model settings from tenant to catalog
- BUG: error processing configuration for chunking patterns in HTML_PROCESSOR
- Removed eveai_chat from docker-files and nginx configuration, as it is now obsolete
- BUG: error in Library Operations when creating a new default RAG library
- BUG: Added public type in migration scripts
- Removed SocketIO from all code and requirements.txt
- eveai_chat becomes deprecated and should be replaced with SSE
- Adaptation of STANDARD_RAG specialist
- Base class definition allowing to realise specialists with crewai framework
- Implementation of SPIN_SPECIALIST
- Implementation of test app for testing specialists (test_specialist_client). Also serves as an example for future SSE-based client
- Improvements to startup scripts to better handle and scale multiple connections
- Small improvements to the interaction forms and views
- Caching implementation improved and augmented with additional caches
- Split differrent caching mechanisms (types, version tree, config) into different cachers
- Improve resource usage on starting components, and correct gevent usage
- Refine repopack usage for eveai_app (too large)
- Change nginx dockerfile to allow for specialist overviews being served statically
- Introduction of dynamic Processors
- Introduction of caching system
- Introduction of a better template manager
- Adaptation of ModelVariables to support dynamic Processors / Retrievers / Specialists
- Start adaptation of chat client
- View License Usages
- Celery Beat container added
- First schedule in Celery Beat for calculating usage (hourly)
- repopack can now split for different components
- Various fixes as consequece of changing file_location / file_name ==> bucket_name / object_name
- Celery Routing / Queuing updated
- Allow filtering on Tenant Types & searching for parts of Tenant names
- Implement health checks
- Start Prometheus monitoring (needs to be finalized)
- Refine audio_processor and srt_processor to reduce duplicate code and support for larger files
- Introduce repopack to reason in LLMs about the code
- Introduction of API-functionality (to be continued). Deduplication of document and url uploads between views and api.
- Improvements on document processing - introduction of processor classes to streamline document inputs
- Removed pure Youtube functionality, as Youtube retrieval of documents continuously changes. But added upload of srt, mp3, ogg and mp4
