Downloads
AI assistant

Help with DAX, Power Pivot and Power BI

The server gives your assistant a searchable knowledge base of DAX, Excel's Data Model (Power Pivot) and Power BI Desktop, plus a formula checker. It uses the same guides you can read in Analyze.

What it knows

How it uses that knowledge

Four tools do the work. You do not call them yourself; the assistant picks them.

search_dax_knowledgeFinds the right section for a question such as "rolling 12 months" or "measure in Power Pivot" and returns working DAX.
get_dax_guideReads a whole topic or one section when the question needs the full picture.
list_dax_topicsLists what is available.
lint_daxChecks a formula for common mistakes: unbalanced brackets, SUM(a*b), a measure inside a Boolean CALCULATE filter, FILTER over a whole table, IFERROR, / instead of DIVIDE.

An example conversation

You: I loaded contoso into Power BI. Give me year-over-year growth and tell me what my date table needs.
Assistant: looks up time intelligence, then answers with a Calendar requirement list (one row per day, no gaps, marked as the date table, related to the fact on its date column) and:
Sales PY = CALCULATE([Total Sales], SAMEPERIODLASTYEAR('Calendar'[Date]))
YoY % = DIVIDE([Total Sales] - [Sales PY], [Sales PY])
and explains that SAMEPERIODLASTYEAR shifts the dates in the current filter context back a year.
You: Check this: Margin % = SUM(Sales[Price] * Sales[Qty]) / SUM(Sales[Cost])
Assistant: runs the checker and replies that SUM() takes one column, so use SUMX(Sales, Sales[Price] * Sales[Qty]), and that DIVIDE is safer than /.

Good questions to ask

What it cannot do

Want a finished example to learn from?

Download a complete Power BI project and ask the assistant to explain a measure in it.