Basedash Models

Product Introduction

1. Definition: Basedash Models is a semantic layer and governed data modeling workspace within the Basedash AI-native business intelligence (BI) platform. It is a technical product category known as a semantic model or a centralized metric layer, built on reusable, governed SQL definitions.
2. Core Value Proposition: It exists to eliminate inconsistent data definitions and metric sprawl across an organization. By providing a single source of truth for core business concepts like customers, orders, and active accounts, it ensures data governance, consistency in reporting, and reliable AI-generated analytics. Its primary value is enabling teams to define metrics once and query them like a table, bridging the gap between complex data infrastructure and business user understanding.

Main Features

1. Semantic Model Workspace: This is not a simple query repository. Each model has dedicated, structured views for Details, Columns, Measures, and Segments. The Details view captures business context, while the Columns view defines the explicit data schema. This structured approach transforms a SQL snippet into a governed business object with clear semantics.
2. Integrated Measures and Segments: Core business metrics (Measures) like MRR or LTV and categorical filters (Segments) like "Enterprise" or "Trialing" are defined and stored directly on the model itself. This prevents metric definitions from fragmenting across individual dashboards or analyst queries, ensuring governed metrics and consistent segmentation for all consumers, whether they are using SQL, charts, or the AI assistant.
3. AI-Native Semantic Integration: The platform's AI assistant is directly integrated with the Models layer. It reads the defined measures, segments, synonyms, row grain, key columns, and usage guidance to generate accurate SQL. This means a business user asking for "active enterprise customers" triggers the assistant to use the pre-defined "Active customers" measure and "Enterprise" segment, ensuring AI-generated SQL aligns with company governance.
4. SQL-Based Abstraction with Table-Like Interface: The technical foundation is governed SQL, serving as the single source of truth. However, users interact with models through a simplified abstraction: SELECT * FROM models.customers. This provides the flexibility and power of SQL with the consistency and reusability of a data model, facilitating SQL-based data modeling that is accessible for analysis.
5. Relationship Mapping and Data Lineage: Models can define explicit relationships to other models (e.g., Customers has many Orders). This enforces join integrity and provides crucial context for both the AI assistant and human analysts, improving data discovery and reducing erroneous joins in complex queries.

Problems Solved

1. Pain Point: Metric and Definition Sprawl. Different departments (sales, finance, marketing) often create their own, slightly varying SQL definitions for "customer," "revenue," or "active user," leading to conflicting reports and decision-making paralysis.
2. Target Audience: Data Teams & Analysts who need to govern and democratize data; Business Operations, Growth, and Finance Teams who consume data for decision-making; Company Leadership requiring a single source of truth for KPIs; Customer-Facing Teams like CS needing reliable customer metrics.
3. Use Cases: Onboarding new team members to a consistent data dictionary; Building company-wide dashboards with certified metrics; Empowering non-technical users to ask the AI assistant complex questions reliably; Auditing and certifying the logic behind key business metrics; Migrating from disparate "definition" queries to a centralized, governed system.

Unique Advantages

1. Differentiation: Unlike traditional BI tools where metrics are defined inside individual dashboard widgets or separate catalog tools, Basedash Models embeds the semantic layer directly into the core analytical workflow. Unlike standalone metric stores, it is natively integrated with an AI assistant and SQL editor, creating a closed-loop system for definition, consumption, and querying.
2. Key Innovation: The deep integration of the semantic model with the generative AI assistant is the key innovation. The assistant acts as a natural language interface to the governed layer, reading pre-defined business logic instead of guessing. This ensures that AI-powered analytics are automatically aligned with company governance, a significant step beyond prompt engineering.

Frequently Asked Questions (FAQ)

1. What is the difference between Basedash Models and traditional BI dimensions/measures? Traditional BI tools often define metrics at the visualization level, leading to duplication. Basedash Models defines them at the data model level as reusable, governed SQL objects, making them available universally across all SQL queries, dashboards, and AI interactions, ensuring a single source of truth.
2. How does Basedash Models handle data governance and access control? Governance is enforced at the model definition stage. By centralizing the logic in a governed workspace, admins control the official definition. The platform's existing role-based access control (RBAC) can manage who can edit models versus who can only query them, securing the semantic layer.
3. Can I use Basedash Models with my existing data warehouse? Yes, Basedash Models sits on top of your existing data warehouse (like Snowflake, BigQuery, or Redshift). It writes and executes governed SQL against your connected data sources. The models.customers reference is an abstraction layer that compiles down to the underlying warehouse SQL.
4. What happens to my old SQL definitions in Basedash? Legacy definitions. references are automatically migrated and remain functional for backward compatibility. The Models workspace is the evolution of that system into a more structured, feature-rich semantic layer.
5. Is the AI assistant required to use Basedash Models? No. The core value of a centralized, queryable semantic layer exists independently. You can query models directly using SQL (SELECT * FROM models.customers). However, the AI assistant leverages the models to provide more accurate, governed answers, creating a powerful synergy.