/analytics
AI Analytics
Site
1 / 8
AI Analytics · natural language → report

Ask a question.
Get the report — fields and all.

Exl Athena turns a plain-English question into a live, governed report. Its AI reads your intent, auto-selects the right fields, assigns each its role, writes the SQL, and renders the result — so anyone can self-serve analytics without knowing a single table or column name.

1
sentence in → a full report out
no drag-and-drop, no field hunting
AI auto-selects the fields
and assigns dimension vs measure
SQL
every answer is transparent
business meaning + the exact query
Ask · in your own words Auto-select · AI picks the fields See · the report + the SQL behind it
Navigate with or the dots below · full-screen for presenting
The problem

“Self-service” BI still needs an analyst.

To build a report, a business user has to know which table holds the data, which columns to pick, which one to group by, which one to sum, and how the tables join. Most don't — so they file a ticket and wait days for a number.

🧭

Field-name maze

Dozens of tables, cryptic column names, hidden joins. Where does “last quarter's revenue by region” even live?

🧱

Drag-and-drop friction

Even with a visual builder, you must know which field is a dimension, which is a measure, and how to aggregate it.

The analyst queue

So the question becomes a ticket. By the time the report lands, the decision has moved on.

Athena removes the middle step — you ask, the AI selects the fields and builds the report.

01 · The headline feature

Auto Select Field — with AI.

Type a question in the field search and press . Athena finds the right columns and drops them onto the canvas — already assigned the correct role.

Extracts intent from your words — no table or column names needed.

Assigns each field a role: group-by dimension vs aggregated measure — and the time grain.

Builds & runs the report instantly; adds any filters it detected.

exl-athena · AI Analytics — Report Builder
Tables & Fields
Total sales by month for last year
📅 order_date
Σ sales_amount
# order_id
A region
A product
A customer
✨ AI auto-selected 2 fields & 1 filter from your question
Dimension order_date · Month Measure SUM(sales_amount) Filter year = last
Total sales by month
02 · Transparent, not a black box

See exactly how it got the answer.

Every AI report opens up: the business meaning in plain language, the exact SQL, and the semantic intent + join path it used. Trust, then apply.

💬

Business Meaning — what the report shows, in plain English.

SQL

SQL Details — the exact, runnable query. Nothing hidden.

🎯

It even detects the subject area and picks LLM or the deterministic SQL engine.

AI SQL Expert
Input prompt
Total sales by month for last year
SQL Engine🎯 Sales
Business MeaningSQL DetailsSemantic
This report totals sales amount for each calendar month over the last twelve months, grouped by month and ordered chronologically — a monthly sales trend.
SELECT toStartOfMonth(order_date) AS month, SUM(sales_amount) AS total_sales FROM fact_sales WHERE order_date >= addYears(today(), -1) GROUP BY month ORDER BY month
03 · Beyond simple aggregates

When it's too complex for a drag-and-drop.

Ratios, differences between two columns, cycle-time date-diffs, real top-N — the visual builder can't express those. Instead of a misleading chart, Athena writes full SQL and runs it.

Σ/Σ

Detects a complex intent and switches to a written query automatically.

📈

Applies trends too — running total, moving average — on the right column.

You still see the SQL and can edit it before running.

AI Analytics — Manual SQL
⚙ Too complex for the visual builder — Athena wrote SQL instead
Input prompt
On-time delivery rate by supplier, last 90 days
🧠 LLM Engine
SELECT supplier, round(100.0 * countIf(on_time = 1) / count(), 1) AS on_time_rate FROM fact_po WHERE issued_date >= addDays(today(), -90) GROUP BY supplier ORDER BY on_time_rate DESC
Trend Running Total → Cumulative Window last 90 days
04 · On-model & reusable

Answers that stay on the rails.

The AI is scoped to governed subject areas, seeded with a Company FAQ, and everyone can save their best questions to a personal library. Repeat questions are cached for instant answers.

🎯

Subject areas keep answers within the right business domain.

📚

Company FAQ + My Library — start from a vetted question, one click.

Cached results return instantly; clear the cache to force a refresh.

AI SQL Expert — ask a question
Company FAQMy Library
Sales Inventory Procurement
Total sales by month for the last year
Top 10 products by revenue this quarter
Revenue by region vs last year
Average order value by customer segment
ModeLLMStandard (SQL Engine)
05 · Ship it

From a question to a dashboard.

Apply the result to the canvas, pin it to a dashboard, and it's live in BI for everyone — the whole loop, from plain English to a shared decision, in under a minute.

Apply Result drops the report onto the builder canvas to refine.

📌

Save to a dashboard — it's instantly shared in Data Intelligence & BI.

Save to Library so the next person just clicks it.

Sales — monthly trend · report
Total sales by month
$4.28M
Total · last 12 mo
+17%
vs prior year
Nov
Peak month
12
Months
The takeaway

Everyone becomes their own analyst.

Ask in plain English, let the AI select the fields and build the report, verify the SQL, and ship it to a dashboard — self-service analytics that's actually self-service.

💬

Ask

A plain-English question — no schema knowledge required.

Auto-select

AI picks the fields, assigns roles, writes the SQL.

📊

Decide

Verify, apply, and pin it live to a dashboard.

Ask your data your first question.