Elsai

Elsai is a complete platform for building AI-powered applications in Python — from simple LLM integrations to complex autonomous agent systems with knowledge graphs.

Products

Elsai Agents

A lightweight, model-agnostic framework for building AI agents. The language model decides what tools to call, in what order — no brittle chains or rigid pipelines.

from elsai import Agent, tool

@tool
def get_weather(city: str) -> str:
    """Get current weather for a city."""
    return f"It's sunny and 72°F in {city}."

agent = Agent(tools=[get_weather])
agent("What's the weather in Tokyo and Paris?")

→ Elsai Agents docs

---

Elsai Core

Production-ready building blocks: LLM connectors, embeddings, vector stores, retrieval, memory, prompts, and document processing.

from elsai_model.bedrock import BedrockConnector
from elsai_vectordb.chromadb import ChromaVectorDb
from elsai_embeddings.azure_openai import AzureOpenAIEmbeddingModel

llm = BedrockConnector(model_id="us.anthropic.claude-3-7-sonnet-20250219-v1:0", ...)
embedding = AzureOpenAIEmbeddingModel(...)
vectordb = ChromaVectorDb(persist_directory="./db")

→ Elsai Core docs

---

Elsai Graph

Extract entities and relationships from text, store them in Neo4j, and answer questions using graph-powered RAG.

from elsai_graph_constructor import GraphConstructor
from elsai_graph_query_rag import GraphRAGEngine

constructor = GraphConstructor(llm=llm)
_, _, graph_data = constructor.construct_graph(chunks=my_chunks)

engine = GraphRAGEngine(neo4j_retriever, hybrid_retriever, llm)
_, _, answer = engine.ask("Who works at Acme Corp?")

→ Elsai Graph docs

---

Architecture overview

Get started

• Quickstart — build your first agent in 5 minutes
• Installation — install packages
• Elsai Agents · Elsai Core · Elsai Graph
• Release Notes