Use @app.graph to connect LangGraph-based agents with real-time UI in Davia.
To add a graph, first install the required packages:
Here’s a quick demo:
Davia supports LangGraph-style AI agents through the @app.graph
decorator.
A graph represents an AI workflow — a sequence of steps connected through logic and memory (called state). It’s useful for building advanced agents like chatbots, assistants, or tool-using flows.
Davia connects directly to the graph you define and builds a UI so users can interact with it — with real-time inputs, state handling, and streaming output.
A graph is an orchestration framework for building controllable AI agents. It enables complex workflows with customizable architectures, long-term memory, and human-in-the-loop capabilities to handle sophisticated tasks reliably.
Davia seamlessly integrates with LangGraph-style AI agents through the @app.graph
decorator. When you expose a graph to Davia:
If your agent is conversational, Davia can generate a chat-style interface to let users interact with it naturally.
Here’s a full example of a simple one-turn chatbot using LangGraph and Davia.
First, install Davia with LangGraph support (requires Python 3.11+):
Then create your agent:
When you run this file, Davia opens the editor at:
You’ll see a simple interface where:
When you expose a LangGraph agent:
MessagesState
) and builds an entire UI based on it,START → node(s) → END
Use @app.graph to connect LangGraph-based agents with real-time UI in Davia.
To add a graph, first install the required packages:
Here’s a quick demo:
Davia supports LangGraph-style AI agents through the @app.graph
decorator.
A graph represents an AI workflow — a sequence of steps connected through logic and memory (called state). It’s useful for building advanced agents like chatbots, assistants, or tool-using flows.
Davia connects directly to the graph you define and builds a UI so users can interact with it — with real-time inputs, state handling, and streaming output.
A graph is an orchestration framework for building controllable AI agents. It enables complex workflows with customizable architectures, long-term memory, and human-in-the-loop capabilities to handle sophisticated tasks reliably.
Davia seamlessly integrates with LangGraph-style AI agents through the @app.graph
decorator. When you expose a graph to Davia:
If your agent is conversational, Davia can generate a chat-style interface to let users interact with it naturally.
Here’s a full example of a simple one-turn chatbot using LangGraph and Davia.
First, install Davia with LangGraph support (requires Python 3.11+):
Then create your agent:
When you run this file, Davia opens the editor at:
You’ll see a simple interface where:
When you expose a LangGraph agent:
MessagesState
) and builds an entire UI based on it,START → node(s) → END