Is Spotonix a text-to-SQL tool?
No. Spotonix uses an LLM to propose an interpretation, but SQL is compiled from a verified BI Algebra plan over governed Segments and Calculations. The LLM does not write the final SQL directly.
Spotonix vs Text-to-SQL
Every text-to-SQL tool starts from scratch with every question. No memory, no reuse, no composability. Spotonix discovers what your team has already built, composes from proven building blocks, and gets smarter with every question.
Why Text-to-SQL Hits a Wall
Text-to-SQL is a map — it shows every street but doesn't know your routes.
Spotonix is GPS with your route history.
Every trip makes the next one faster. Every question adds to the context.
Head to Head
"Which stores are losing habitual buying customers?"
The Architecture
Parse the question, classify intent, ground every term against your Context Graph.
Resolve interpretations against existing Segments, Calculations, and Analysis Patterns. Compile SQL only from what closes.
Every verified answer persists as reusable context. The next question is faster than the last.
FAQ
No. Spotonix uses an LLM to propose an interpretation, but SQL is compiled from a verified BI Algebra plan over governed Segments and Calculations. The LLM does not write the final SQL directly.
A verified plan is inspectable, reusable, and deterministic. Same accepted plan. Same SQL. Every time. If the plan does not close, Spotonix clarifies or refuses instead of inventing an answer.
Spotonix grounds each question in the customer Context Graph, reuses accepted business logic, shows the interpretation before execution, and preserves accepted answers as reusable context.