Valenova

About the Project

Valenova is an enterprise AI adoption platform delivered as both a cross-platform desktop app (React + Electron + Express.js local server) and a Next.js marketing website deployed on Vercel. Supabase handles authentication (Google OAuth), user data, and zero-knowledge query metadata. A custom Model Context Protocol (MCP) server bridges the AI agent layer to enterprise databases, Oracle, SAP, PostgreSQL, MySQL, Snowflake, without raw data ever leaving the client's infrastructure. LemonSqueezy powers license-based billing; the AI agent orchestrates multi-hop text-to-SQL query planning across federated data sources in under a second.

How It Works

1. 1

   The Electron desktop app bundles a React frontend and an Express.js local server, the Express server runs the custom MCP that connects to the enterprise's local databases via read-only JDBC/ODBC drivers, keeping raw data entirely within the client's environment.

2. 2

   Google OAuth via Supabase authenticates users across both the desktop app and the Next.js web dashboard; Supabase stores only encrypted query plans and aggregated result metadata, never raw records.

3. 3

   The custom MCP server exposes a standardised tool interface that the AI agent layer calls to introspect schema metadata, execute sub-queries, and federate results across multiple data sources in a single natural language request.

4. 4

   The AI agent decomposes multi-hop questions into sub-queries, routes each to the correct data source via the MCP, federates results in-memory, and synthesises a response as a table, chart, or prose depending on query intent.

5. 5

   LemonSqueezy handles perpetual and subscription license billing for the desktop app with automatic entitlement checks against Supabase on startup; the Next.js website on Vercel serves as the marketing, documentation, and account management hub.

Tech Stack