Appearance
elsai Core v1.0
elsai Core is a modular framework of Python packages for building LLM-powered applications — models, embeddings, vector stores, retrievers, document processing, and integrations.
Packages in v1.0
| Package | Version | Purpose |
|---|---|---|
elsai-model | 1.0.0 | LLM connectors — OpenAI, Azure OpenAI, AWS Bedrock |
elsai-embeddings | 0.1.0 | Azure OpenAI embeddings |
elsai-vectordb | 1.1.0 | ChromaDB, Pinecone, Weaviate |
elsai-retrievers | 0.1.0 | Hybrid retrieval (semantic + BM25) |
elsai-chat-history | — | Conversation history management |
elsai-prompts | 1.0.0 | Pezzo-based prompt management |
elsai-utilities | 0.1.0 | Document chunking and conversion |
elsai-text-extractors | 0.1.0 | PDF, DOCX, CSV, Excel extraction |
elsai-ocr-extractors | 1.0.0 | Azure, LlamaParse, VisionAI OCR |
elsai-cloud-connectors | 0.1.0 | S3, Azure Blob, SharePoint, OneDrive |
elsai-db | 0.1.0 | Natural language queries over SQL |
elsai-stt | 0.1.0 | Azure Whisper STT / TTS |
elsai-mcp | 0.1.0 | Swagger → MCP server |
INFO
Looking for the latest version? Switch to v3.0 which introduces the unified *Model API. For v2.0 connector docs see v2.0 legacy.