Guides for verifiable AI analytics
Practical guides for business users and analysts: query databases in natural language, review the SQL behind an answer, build reusable reports and compare analytics tools. Explore worked examples, connector requirements and read-only controls.
New here? Start with what it means to chat with your database in plain English, then follow the guide it points you to.
All guides
Getting answers
- Technical guide
Ask MySQL Questions in Plain English
Build MySQL reports with Datarelix: define an order question, obtain read access, check authentication and replica freshness, and review the generated SQL.
- Technical guide
Query BigQuery in Plain English
Query BigQuery in plain language with Datarelix. Set the reporting period, check dataset access and location, and understand partition scans and query limits.
- Technical guide
Ask PostgreSQL Questions in Plain English
Ask PostgreSQL reporting questions with Datarelix. Learn what access you need, how to check a sales result, and when to ask the database owner for help.
- Technical guide
Ask Snowflake Questions in Plain English
Explore Snowflake reporting with Datarelix: business questions, authentication prerequisites, warehouse access, query review and compute costs.
- Guide
Why Your AI Tool Picks the Wrong Column
Improve AI data questions with useful table and column descriptions. Explain metrics, statuses and joins, then compare results with known answers.
- Guide
Text-to-SQL Explained: How Natural Language Becomes a Database Query
Understand text-to-SQL with a reporting example: how a question becomes SQL, what schema context supplies, and what to check before using the result.
- Guide
How to Verify an AI-Generated Data Answer Before It Reaches the Board
Check AI-generated reports: business definitions, dates, sources, joins, missing records, and when to ask an analyst to review the SQL.
- Guide
What Is Verifiable Conversational Analytics?
Learn what conversational analytics shows behind an answer: executed queries, source tables, returned rows, and checks to make before using a number.
Reporting and dashboards
- Guide
Forecast Monthly Sales From Your Database
Build a monthly sales forecast in Datarelix: define net sales, check historical coverage, compare past tests and choose how to refresh the result.
- Guide
Introducing Forecasting Pipelines in Datarelix
Datarelix forecasting is in beta: describe a business measure, review the data and model comparison, then publish and refresh your forecast.
- Guide
Build a dashboard from database questions
Build a Datarelix dashboard from database questions: review a result, create a chart, save it and refresh manually, with access and sharing limits explained.
- Guide
Find delayed orders in your operations data
Find delayed orders in your database: define the cutoff, select status and dispatch timestamps, compare periods and check an actionable result in Datarelix.
- Guide
Create a weekly sales report from your database
Create a weekly database sales report with clear revenue and refund definitions, customer segments, date boundaries and reusable charts in Datarelix.
- Guide
Build Self-Service Analytics Without Replacing Your Data Stack
Start self-service reporting with a database you already use: get appropriate access, define a first question, check the result, and save useful charts.
Choosing tools
- Guide
Text-to-SQL vs writing SQL manually: when to use each
Compare text-to-SQL and manual SQL on the same sales report: define the question, generate a query, check results, revise filters and reuse the answer.
- Guide
AI Analytics vs Traditional BI: When to Use Each
Compare conversational analytics and BI for ad hoc questions, recurring reports and shared metrics. Choose by workflow, access needs and how results are reviewed.
Access and security
- Security & trust
AI Analytics Security Checklist: 10 Vendor Questions
A practical AI analytics security checklist covering access, model inputs, saved results, query controls and requirements specific to your organisation.
- Security & trust
How Datarelix Keeps Database Credentials Out of the AI Model
Understand how Datarelix handles database credentials, schema metadata, stored query results and values used in follow-up AI questions.
- Security & trust
Read-Only AI Analytics: Architecture, Controls, and Threat Model
Evaluate read-only AI analytics: database permissions, query validation, generated-code controls, and the risks that still need review.
Beyond the guides
The security page documents the enforced boundaries these guides describe, and the documentation covers connection setup for every supported engine.