The Colrows AI Data Analyst is an intelligent agent that enables natural-language analytics across your connected data sources.

It understands your business context, translates questions into optimized SQL, and generates visual insights automatically — without manual query writing or BI modeling.

Backed by Colrows’ Semantic Layer (aka SemantIQ), it reasons over business terms, metrics, and data relationships to produce accurate, explainable answers.

Key Highlights

• Multi-source SQL generation: Generates dialect-specific queries across PostgreSQL, Snowflake, BigQuery, Redshift, MySQL, and others.
• Semantic Grounding: Uses SemantIQ graphs to resolve business terms, metrics, tables, and relationships.
• Data Retrieval: Retrieves data from a variety of data sources and allows exploration and export of it.
• Autonomous Charting: Suggests and renders the most appropriate chart automatically.
• Team collaboration: Save, share, or publish results as charts or dashboards.
• Explainable AI: Displays generated SQL and lineage for transparency and debugging.

When a user asks a question in plain English — for example,

“Show revenue growth by region for the last two quarters.”

The AI Data Analyst performs a multi-stage reasoning workflow:

1. Semantic Understanding

   • Parses the question and resolves business terms and metrics from the Colrows Semantic Layer.
   • Identifies relevant datasets, columns, filters, and relationships.

2. SQL Generation and Optimization

   • Constructs vendor-specific SQL for the connected source (e.g., Snowflake, PostgreSQL, BigQuery).
   • Applies optimization rules and access policies enforced by the Colrows Platform.

3. Validation

   • Validates the query for syntax and semantic correctness
   • Reconciles with Knowledge Graph
   • Runs coverage checks
   • Verifies cardinality and runs distinct value checks

4. Execution and Visualization

   • Executes the generated SQL via the Colrows query engine.
   • Automatically selects suitable chart types (e.g., trend, distribution, comparison) and renders interactive visuals.

5. Explainability and Learning

   • Displays the generated SQL and reasoning path for transparency.
   • Learns from user feedback to refine metric definitions and semantic understanding.

Colrows enforces strict data-governance principles:

1. No Data Storage: Query results are transient; Colrows never stores customer data.

2. Access Policy Enforcement: SQL is executed only with the authenticated user’s credentials and governed by role-based policies.

3. Explainable AI: Every response is accompanied by the generated SQL and schema lineage for traceability.

Natural Language Querying

Ask questions in plain English, such as:

• “Top 10 customers by revenue this quarter”
• “Compare conversion rate across campaigns”
• “How many users upgraded after trial?”

The AI Data Analyst interprets business intent, maps it to the semantic layer, and retrieves precise answers — no SQL required.

Autonomous SQL Generation

Colrows automatically generates optimized, dialect-specific SQL for each connected source.

The engine ensures:

• Correct schema resolution
• Automatic join detection
• Filter and aggregation inference
• Compliance with user-level access policies

Generated SQL can be reviewed, modified, or executed directly.

AI-Generated Charts

Each query result is visualized instantly with AI-selected chart types — bar, line, trend, table, or distribution plots — chosen based on data shape and metric semantics.

Users can:

• Switch visualization types
• Customize labels and legends
• Save and share charts within the workspace
• Export to image or CSV

Dashboards

Multiple charts can be combined into a single interactive dashboard.

Dashboards support:

• Cross-filtering between charts
• Dynamic resizing and layout editing
• AI-suggested summaries and KPI highlights
• Scheduled refreshes and live updates from data sources

Chart and Dashboard Sharing

Colrows supports multi-level sharing for collaboration and governed visibility:

• Internal Sharing: Charts and dashboards can be shared with individual users or groups within the organization. Permissions determine view or edit access.
• Public Links: Users can generate secure, time-bound public links to share insights outside the organization (e.g., clients, partners, or presentations).
• Access Auditing: All shares are logged with the creator, recipient, and timestamp for compliance visibility.

This enables teams to collaborate fluidly while maintaining enterprise-level governance.

Conversation and History Retention

Colrows retains each user’s AI chat history — including all questions, generated SQL, tables, and charts — creating a persistent analytical memory.

• Searchable History: Users can revisit any past conversation, chart, or table.
• Live Refresh: When a history item is opened, both the chat context and data tables refresh in real time to reflect the latest underlying data.
• Context Continuity: The AI Data Analyst remembers previous questions within the same conversation, enabling follow-up queries such as “compare this to last quarter.”

This conversational persistence allows analysts to build cumulative understanding over time without re-entering queries.

Colrows AI Data Analyst supports any data store with SQL interface, including all major relational databases, ensuring seamless connectivity, data querying, and SQL dialect optimization.

Each connector is natively integrated with the Colrows Platform and supports full semantic resolution, governed query execution, and real-time visualization.

Analytical Databases and Warehouses

• Snowflake
• Google BigQuery
• Amazon Redshift
• Databricks SQL
• Azure Synapse Analytics
• ClickHouse
• Trino
• PrestoDB / PrestoSQL

Relational Databases

• PostgreSQL
• MySQL
• MariaDB
• Microsoft SQL Server
• Oracle Database
• IBM DB2
• SAP HANA