Quick Start

This guide will help you get started with using an existing DataBridge server. If you need to set up your own server, see the Installation Guide instead.

Getting Your Access URI

Note: For local development, you can skip this step and connect directly to http://localhost:8000 without authentication.

If you need authentication:

1. Visit your DataBridge server's API documentation (http://localhost:8000/docs)

2. Find and use the /local/generate_uri endpoint to generate your URI

3. Save this URI - you'll need it to connect to the server with authentication

Ways to Use DataBridge

You can interact with DataBridge in several ways:

1. Using the Shell

The shell provides an interactive Python environment for quick testing:

# Without authentication (connects to localhost):
python shell.py

# With authentication (using your generated URI):
python shell.py "databridge://user:token@localhost:8000"

Once connected, you can interactively run commands:

>>> db.ingest_text("Machine learning is transforming industries...", 
...                metadata={"title": "ML Overview"})
{'external_id': 'doc_123', 'metadata': {'title': 'ML Overview'}, ...}

>>> results = db.retrieve_chunks("What are the key findings?")
>>> for r in results:
...     print(f"Score: {r['score']}")
...     print(f"Content: {r['content']}\n")
Score: 0.89
Content: Machine learning is transforming industries...

The shell supports all SDK functionality and allows you to write multi-line code just like in a Python script.

2. Using the Python SDK

# Without authentication (connects to localhost):
db = DataBridge()

# With authentication (using your generated URI):
db = DataBridge("databridge://user:token@localhost:8000")

# Example usage:
doc = db.ingest_text(
    content="Machine learning is transforming industries...",
    metadata={"title": "ML Overview"}
)

3. Using the UI Component

The UI component provides a convenient interface for prototyping and testing:

1. Access the UI at http://localhost:3000

2. For authenticated access, enter your generated URI in the connection field

3. For local development, just click Connect (no URI needed)

The UI lets you:

• Test different document types and upload workflows

• Experiment with various query prompts

• Monitor document processing and indexing

• Debug and visualize search results

Next Steps

Learn more about:

• API Reference

• Python SDK

• Ingestion

• Querying