Smart Analytics & Generative Engine, an enterprise solution based on open source

This is the Advanced Analytics platform that combines the power of enterprise tools with the freedom of open source. Built on Apache Superset and enhanced with a proprietary semantic layer and AI agents, it delivers self-service analytics with no licensing constraints and no vendor lock-in. Fully customizable to client needs, it adapts to any business domain and existing infrastructure.

Scenario

Organizations tackling data analytics initiatives face a recurring dilemma:

  • Enterprise tools (Power BI, Tableau, Qlik) offer mature governance and a polished UX, but impose high per-user costs, vendor dependency and limited portability

  • Open-source solutions (Apache Superset, Metabase, Redash) provide flexibility and zero lock-in, but fall short on semantic layer capabilities, AI assistance, and self-service experience for non-technical users

The outcome? Data teams are forced to choose between analytics quality and economic sustainability.

Benefits

  • Zero Vendor Lock-in: Data, infrastructure, and code remain the property of the client. No binding contract.

  • AI-First by design: Creating charts in natural language and AI-assisted modeling are integrated features, not paid add-ons.

  • No user costs: Costs scale with infrastructure, not with the number of analysts. Users can be added without increasing the budget.

  • Versionable Semantic Layer: The file model.yml is versioned in Git, transparent, traceable, and modifiable independently.

  • Open and established standards: Apache Superset is the most widely used open-source BI platform in the world, supported by an active community.

  • Progressive scalability: Deployable on any infrastructure, from a single host to a Kubernetes cluster, for a level of scalability that grows with the organization's needs, without platform migrations.

  • Fully customizable: Every component (semantic model, dashboard, AI agents) is configurable and adaptable to the client's specific processes, terminology, and KPIs.

Solution

These results are achieved through three in-house developed tools that cover the entire journey from semantic modeling to visualization. Each tool is designed for a specific profile (data engineer, analyst, business user) and works on the same shared model

Models Builder — build your semantic layer with AI

The starting point of every project. Models Builder is the AI agent that turns a raw database schema into a structured semantic model: it scans tables automatically, proposes relationships and business-friendly names and it always waits for human validation before applying any change. The result is a Git-versionable YAML file — transparent, auditable and fully owned by the client.

Datasets Builder — select exactly the data you need

Based on the semantic model, Datasets Builder lets you compose virtual datasets by selecting the tables and columns relevant to a specific analytical domain. The tool validates connectivity between selected entities in real time and generates the dataset ready for exploration in Superset. No code, no data duplication.

Charts Builder — describe the chart, the AI builds it

The user describes the chart they want to see in natural language. Charts Builder interprets the intent, anchors it to the semantic model and returns a validated chart specification, refinable through conversational iteration. From brief to published chart, without going through the traditional exploration interface.

Costs

No per-seat licensing: costs scale with infrastructure, not with the number of analysts. The pricing model consists of an initial setup (provisioning, deployment and the first semantic model) and a recurring fee tied to cloud infrastructure and application support. Adding users doesn't add budget.