Disclaimers

If the user prompt does not specify a timeframe, PowerBuddy defaults to the current academic year. If the current date falls outside the defined start and end dates of the school year, it defaults to the next academic year.

• When no records are returned – likely because the next academic year has not yet started, PowerBuddy displays suggested questions along with the assumptions.

Questions that fall outside the supported scope may result in an error on the UI.

AI rephrased query functionality that summarizes the context from follow-up questions will be added back in the future release.

Some analytical queries may produce slightly different results based on the filters and groupings:
Example – Show YoY percentage change in attendance rates for homeless students over the last 3 years, grouped by school type.
If a student transfers between schools during the year, they might be counted more than once depending on how the data is grouped. This can lead to slight differences in the percentages when certain filters (like school number) are included or excluded.

‘Reports’ tab has a maximum threshold of processing up to a maximum of 50,000 records to produce charts and summary. Data dashboard provides a preview of up to 500 records in the UI. As a result, counts or summaries may differ slightly between the Reports and Data Dashboard views

You are responsible for the accuracy and appropriateness of the queries you submit. PowerBuddy processes user input but does not validate the underlying assumptions or context.

PowerBuddy has access to the entire underlying data and should only be accessed by users who have the appropriate rights to see that data in accordance with FERPA and any other data privacy laws.

While the GPT model used in PowerBuddy for Data Analysis is powerful, it may not always understand nuanced or highly specialized queries. If the initial response is not what you expected, consider rephrasing your query, providing specific instructions and try again. Users are encouraged to submit their feedback with the thumbs up /down options in the UI

AI may not have a complete domain knowledge of all the tables in a product, so it depends on you to ask questions clearly.

PowerBuddy cannot interpret charts and, therefore, cannot provide suggestions derived from charts. It can only offer analysis derived from the underlying data tables.

AI may join tables in in ways that seem unclear. This can happen because it is trying to guess how to best filter the data or because of confusing instructions. AI is not perfect and can make mistakes, too. if the response does not meet your expectations, rephrase your query, provide specific instructions and try again.

Any query that is written in the Enter a SQL Query field is across all domains and tables within the selected product, to which all users have read-only access in PowerBuddy for Data Analysis.

Processing large datasets can result in slow performance.

PowerBuddy for Data Analysis is capable of basic regression, but the quality of the result depends on chosen variables and instructions. It may produce unexpected responses, and you should interpret results with awareness of these limitations.

Currently, PowerBuddy for Data Analysis lacks knowledge of customer-defined conversion, calculations or tables. These will be addressed in the future releases.

Certain business logic like GPA calculations, graduation requirements are not currently supported since they are usually conditional on many factors and vary considerably from school to school.