Answer Preparation Steps

This section describes how to use the EXPLAIN tab under Query Inspector.

After you ask a question, Connecty shows a real-time thinking process.

Its follows deep reasoning powered four-step process — Intent → Context → Grammar → Query — to ensure that each question is correctly understood, mapped to the right data, grounded in your organization's metric definitions and translated into precise analytical output.

Each step represents a specific layer of understanding:

• Intent determines what the user is asking.

• Context identifies where in the data the answer can be found.

• Grammar defines how that logic is expressed syntactically.

• Query produces the final SQL that retrieves the result.

Each step displays its execution time, expand/collapse preview, and quick-access buttons for copying code, opening visual graphs and adding instructions (natural-language commands for precision editing or SmartNode Editor).

1. Intent

The Intent stage identifies the analytical goal behind a user’s question. It determines what the user wants to measure, compare, or understand and defines the main objective of the query before any database operations are applied.

2. Context

The Context stage determines where and how the identified intent is executed across connected data sources. At this step, Connecty AI maps the intent to the relevant tables, fields and relationships defined in the Metricverse.

Deep reasoning is applied to identify the relevant context. For actionable explainability, the reasoning result is exposed to the user with confidence scores.

By analyzing these objects and metric entities together, Connecty AI constructs a semantic map of the data - determining how each table connects through primary and foreign keys. This ensures that the generated SQL query joins all necessary entities to deliver an accurate and context-aware answer.

3. Grammar

The Grammar view presents this logic in a graph based layout, where each block corresponds to a SQL operation — from selecting and filtering fields to applying aggregations. Examples include nodes such as ‘Supplier Name’, ‘Region Name’, ‘Asia Region Only’, and ‘Average Total Price’. The connecting lines show how tables and expressions flow together, forming the complete logical structure of the final SQL query.

4. Query

The Query stage is where Connecty AI converts the structured grammar into executable SQL and runs it against the connected data source. This step produces the final result — the generated SQL statement.

User can use AI SQL Editor to edit the query using conversational analytics, or manually modifying the SQL.