Natural language analytics,
on your own infrastructure
Savvina AI turns plain-English questions into precise SQL — instantly, privately, and entirely within your own environment. It connects to your databases, generates queries using the LLM provider of your choice, and returns results without any data leaving your network.
We built it because organisations that work with sensitive data — financial services, healthcare, HR, logistics — should not have to choose between powerful analytics and keeping their data in-house. Self-hosting should be the easy option, not the hard one.
Design principles
Adapter pattern everywhere
Every data source and every LLM provider is a plugin. Adding a new one requires a single Python file — no changes to any existing code. This is the architectural core that makes Savvina AI extensible without fragility.
Privacy by design
Query results never leave your network. The LLM only ever sees schema metadata — table names, column types, optional sample values. Per-connection controls let you exclude entire schemas, tables, or columns from the LLM prompt entirely.
Zero-change extensibility
New datasource adapters and LLM providers register themselves via a decorator pattern at import time. The core NL-to-SQL pipeline requires no modification to support a new target — it discovers registered adapters automatically on startup.
No proprietary lock-in
Savvina AI Community Edition is open-source under BSL 1.1, converting to Apache 2.0 on 2030-06-01. No license key, no usage limits, no phone-home. You can inspect every line of code that processes your data.
Technology stack
Licensing
Business Source License 1.1 → Apache 2.0
The Community Edition is distributed under the Business Source License 1.1 (BSL 1.1). It is free for development, testing, evaluation, personal projects, and non-commercial internal use — with no license key, no feature restrictions, and no phone-home.
On 2030-06-01 the license converts automatically to the Apache License 2.0, removing all restrictions.
A commercial license is required if you:
- Deploy Savvina AI in a production environment for business purposes
- Offer it as a hosted or managed service to third parties
- Embed it into a product or SaaS platform you sell or charge for
Contact info@savvina.ai for commercial licensing.
Get in touch
Questions about Savvina AI, enterprise deployments, or partnerships? Email us at info@savvina.ai or visit the contact page.