v0.1.6

##Frontend

###Added

Added the resolved entities modal to view the full answer with resolved entities
Added a rules panel to the UI to view current entity resolution rules
Can undo rules in the UI

###Changed

Allow query without source data if mention is present
Added resolve entity rule

##Backend

###Added

Added support for queries without source data in vector database
Graceful failure of triple export when no chunks are found
Tested Qdrant vector database service
Added resolve entity rule

###Changed

Separated embedding service from LLM service

v0.1.5

Added

• Updating backend to work with new UI

• UI: Global rules

• UI: Reactgrid integration

• UI: "Replace" rule

• UI: Multi-table support

• Changed

• Updating backend to work with new UI

• Tweaked query output for separating answers and chunks

• Support for Qdrant vector database.

• Updated Milvus reference in the factory to be more robust to other Milvus datastores

Improved

• Updating testing to Mock OpenAI client and embeddings

v0.1.4

Frontend:

• column mention highlighting was added, and csv + export triples were refactored

Backend:

• updated test files to be less restrictive
• refactored factories and services to be more extensible

v0.1.3

Changed

• Restructured project directory for improved organization and scalability
• Moved router files to new api/v1/endpoints directory
• Created new core directory for fundamental application components
• Updated llm operations to new services in services/llm
• Updated vector database operations to new services in services/vector_db

Improved

• Separated configuration from dependency injection for better maintainability and clarity
• Created new utils directory for fundamental application components
• Seperated Pydantic models into models and schemas directories

v0.1.2

Separating LLM concerns and modularising generated retrieval

v0.1.1

Integrating Unstructured API for document parsing

v0.1.0

Initial commit