Knowledge Overview

Knowledge is where you capture trusted, reviewed context that guides how NQRust Analytics writes SQL. By recording the way your organization defines metrics and the rules your queries should follow, you give the system reliable ground truth to draw on rather than leaving it to guess. It is most valuable for domain-specific logic and for questions your team asks repeatedly.

NQRust Analytics supports two kinds of knowledge.

Question–SQL pairs

A question–SQL pair links a natural-language question to a verified SQL query. Once a pair is saved, future questions that resemble it can reuse that proven SQL as a reference, which keeps answers accurate and consistent over time.

Use question–SQL pairs when:

• A metric must be calculated with a precise, agreed-upon methodology.
• Business logic is complex enough that the exact SQL formulation matters.
• The same business questions come up regularly with an established answer.
• A KPI is specific to your organization and not obvious from the schema alone.
• A query is error-prone because it joins across many tables.

For example, if your team computes "customer lifetime value" using a particular method, you can store the question "What is our average customer lifetime value?" alongside the SQL that calculates it correctly.

Learn more in Question–SQL pairs.

Instructions

Instructions are guidelines that shape how SQL is generated. They take two forms:

• Global instructions apply to every query and express general rules about your data model and business.
• Question-matching instructions apply only when a question fits a pattern or topic you define.

Instructions help the system understand your data model and business context up front, so it produces correct SQL with less back-and-forth refinement.

Learn more in Instructions.

Why knowledge matters

• Consistency — similar questions resolve to your verified SQL, so different users get the same reliable answer.
• Accuracy — pre-verified SQL reduces mistakes on complex queries.
• Domain expertise — your organization's calculation methods and rules are captured and reused.

Knowledge is one of the layers that turns a schema into real understanding of your business. See What is context? for how the pieces fit together.