Embedding Service#1286

Draft

EmanueleDeRossi1 wants to merge 26 commits intomainfrom

feature/embedding-service

Commits on Feb 13, 2026

feat: consistently use 'language model' instead of 'llm model'
EmanueleDeRossi1
committed

Commits on Feb 14, 2026

feat: standardize terminology to 'language model' and 'embedding model'
EmanueleDeRossi1
committed

Commits on Feb 16, 2026

feat: use consistent naming for language and embedding model in tests and backend
EmanueleDeRossi1
committed
feat(sdk): get_model to get_language_model for clarity
EmanueleDeRossi1
committed
feat(test): update tests for renaming in SDK
EmanueleDeRossi1
committed
refactor: rename model config vars for clarity
EmanueleDeRossi1
committed
rename model_type to purpose
EmanueleDeRossi1
committed
refactor(frontend): change model_type from llm -> language
EmanueleDeRossi1
committed
refactor(sdk): change llm to language in model_type param
EmanueleDeRossi1
committed
feat: add default rhesis embedding model support
EmanueleDeRossi1
committed
change from 'language_model' to 'model' in schemas, routers and model connection
EmanueleDeRossi1
committed
fix: import in tests and remove unused aliases
EmanueleDeRossi1
committed
fix(test): use rhesis default embedding model
EmanueleDeRossi1
committed
fix: import
EmanueleDeRossi1
committed
fix(test): 'get_model'-> 'get_language_model'
EmanueleDeRossi1
committed
feat: add multi-dimension embedding storage system
EmanueleDeRossi1
committed
feat(db): enable pgvector extension for vector storage
EmanueleDeRossi1
committed
feat: enhance embedding table with multi-dimension support and full-text search
EmanueleDeRossi1
committed
feat: add EmbeddableMixin for searchable text generation
EmanueleDeRossi1
committed
fix: remove ABC inheritance from EmbeddableMixin
EmanueleDeRossi1
committed
feat(services): add AsyncService base class for async/sync task orchestration
EmanueleDeRossi1
committed
refactor(telemetry): migrate EnrichmentService to AsyncService base
EmanueleDeRossi1
committed
feat(models): add dimension auto-detection for embedding models
EmanueleDeRossi1
committed
feat(embedding): add unique constraint to embedding model to avoid multiple identical active embeddings
EmanueleDeRossi1
committed
refactor: simplify embedding connection test by removing unused endpoint param
EmanueleDeRossi1
committed
feat: add EmbeddingGenerator for entity embedding creation
EmanueleDeRossi1
committed