Reasoning you can read
Each answer lists the steps taken, from looking up the definition to running the query and checking the result.
Explainable AI
Answers that show their work.
Every Chat4ED answer comes with the reasoning behind it: the business definitions it applied, the SQL it ran, the tables it read and the checks it passed. Your teams and your auditors can see exactly where a number came from.
The problem
A black-box answer is useless in a regulated business. If a CFO cannot show an auditor how a figure was produced, the figure cannot be used, however quickly it arrived.
SQL and sources attached
Open "Why this answer?" to see the exact SQL, the tables and columns used, and the business definitions applied.
A full audit trail
Every question, query and answer is logged with who asked it, so any number can be traced back later.
Sample answer
What an answer looks like.
Illustrative example on sample data.
Where did we miss the Q3 opex budget, and what happens to Q4 if nothing changes?
Answer 3 of 3 models agreeQ3 opex was $1.3M (6.1%) over budget. Marketing (+$0.7M, events) and Operations (+$0.5M, overtime) caused 92% of the miss. At the current run-rate, Q4 lands $1.6M over.
SQL it ran
SELECT d.department,
SUM(b.amount) AS budget,
SUM(a.amount) AS actual
FROM finance.opex_actuals a
JOIN finance.opex_budget b ON a.cost_center = b.cost_center
JOIN org.departments d ON d.id = a.department_id
WHERE a.fiscal_quarter = '2026-Q3'
GROUP BY d.department;Definitions used
- Opex excludes COGS and one-off restructuring
Sources
finance.opex_actualsfinance.opex_budgetTry asking
Questions teams ask.
“Why is Q3 opex over budget?”
The departments behind the miss, how much each contributed, and the SQL that compared budget to actuals.
“How did you calculate churn?”
The churn definition used, the time window and the tables behind it.
“Which data did this forecast use?”
The source tables, the date range and the backtest error of the model.
Connectors
Works with the data you already have.
Snowflake
BigQuery
Redshift
PostgreSQL
MySQL
SQL Server
Oracle
ClickHouse
Trino- Athena
SAP HANA & B1
Excel & CSV