Vanna

What is Vanna

Vanna is an AI-powered tool that allows users to interact with their databases by generating SQL queries through natural language queries, thus reducing the need for SQL expertise. It supports many databases, ensures high accuracy through training data, and prioritizes security by minimizing data exposure. Vanna's open-source nature and customizable front ends make it adaptable for various business needs.

Key Features of Vanna

- Open-Source Platform: The Vanna Python package and various frontend integrations are open-source, allowing you to run Vanna on your own infrastructure.

- High Accuracy on Complex Datasets: Vanna’s accuracy improves with the volume and complexity of the training data you supply, making it well-suited for large datasets.

- Designed for Security: Database contents are never sent to the LLM unless specifically enabled, with the metadata storage layer only seeing schemas, documentation, and queries.

- Self-Learning Capability: The model continuously improves as it augments its training data through regular use.

- Supports Many Databases: Out-of-the-box support for Snowflake, BigQuery, Postgres, and many other databases, with easy configurability for additional connectors.

- Flexible Frontend Options: Start in a Jupyter Notebook, expose to business users via Slackbot, web app, or Streamlit app, and integrate into your own web app if desired.

- Fully Customizable: Tailor the platform to meet your unique needs, with options for enterprise-level customization and deployment.

- Quick Insights Generation: Generate SQL queries for any SQL database by simply asking questions, saving time and increasing productivity.

- Enterprise Customization Support: Large enterprises can customize deployment themselves or seek help from experienced developers through Vanna’s network.

Pricing

Community Edition:

• Cost: Free
  • Open source
• Features:
  • Use any LLM
  • Use any vector database
  • Discord community support

Premium Plan:

• Cost: Not specified (monthly)
  • Good for scaling
• Features:
  • GPT-4 class LLM
  • 500 LLM requests per month
  • Hosted metadata retrieval
  • E-mail support
  • Additional Services:
  • Onboarding support: $100 one-time fee
  • Extra 1000 requests: $50 monthly

Enterprise Plan:

• Cost: Contact for pricing
  • Full customization
• Features:
  • Meet any InfoSec requirements
  • On-prem deployment
  • Priority support