How the Spotonix Analyst thinks.

Each moment below is a real question from a real evaluation run. Each shows one thing Spotonix does that an LLM-only BI tool structurally can't.

01 / Learns Your Business

Learns what your team already built.

Schema, dbt models, BI assets, every SQL query your team has ever run. Spotonix studies the corpus and constructs a Context Graph — yours alone.

Your stack:
Warehouse Snowflake schema tables · columns · types
Transformations dbt models refs · sources · materialized
BI assets Power BI · DAX models measures · hierarchies · relationships
Activity SQL query history 12 months · all users

Your Context Graph · TPC-DS retail

Customers Segment Premium Customers Segment High Spenders Segment Stores Segment Sales Calc · conformed Categories Dimension Dimension · Channel Calc · QoQ Retention Calc · YoY Growth Premium Count Answer · 1,021
Segments Calculations Answers Dimensions
18,432 SQL queries analyzed
142 dashboards parsed
89 dbt models read
47 building blocks composed

02 / Clarify once · Memory forever

When a term is ambiguous, it asks. Once.

First time Spotonix sees an ambiguous term, it asks you to clarify or rephrase. Every subsequent question uses your validated definition automatically — no re-asking. The clarification is a one-time tax. The memory is permanent.

Turn 1 First time it sees "high spenders"
> What Product Categories do High Spenders like to shop?
?

"High spenders" matches two Segments. Pick one — or rephrase.

"Spend > $5,000 / year" validated and added to your Context Graph
Turn 2 Any subsequent question
> List of Top 5 Categories of products high spenders are buying this year.

Plan · auto-applied from Context Graph

"high spenders" Spend > $5,000 / year Segment · from your Context Graph · last validated 6 days ago
"Top 5 Categories" rank(sum(sales)) desc, limit 5 Calculation · BI Algebra: ranking
"this year" year = current_year Dimension · time filter
No clarification needed. Algebra closed · SQL compiled · 5 categories returned

Clarification once. Memory forever. Every disambiguation strengthens the Context Graph — and shortens the next question.

03 / Interpret

Read the plan before any SQL runs.

A real, multi-concept question. Every interpretation is a graph — named nodes, named edges — that you can read in ten seconds. The plan is the trust mechanism. SQL only compiles after the algebra closes.

> Which stores are losing habitual buying customers over the last 4 quarters?
Answer Stores with Declining Habitual Buyers 12 stores · ranked by QoQ drop
resolves to
Segment · found Habitual Buyers customers with ≥ 4 purchases per quarter · spend ≥ $500
grounded on
Calculation · BI Algebra: growth QoQ Customer Count Change (current − prior) / prior · threshold < 0
scoped by
Dimensions Store × Quarter cross-tabulated over the last 4 quarters
Algebra closes · plan accepted · SQL compiled deterministically in 2.1s same plan → same SQL, forever

04 / Conformed measures

Three fact tables. One Sales measure.

"Sales" lives in three places in your warehouse — store, catalog, web — under three different column names. Building this reconciliation in a semantic layer typically takes weeks. Spotonix does it before the first question.

> Show me the YoY Sales percentage change.
Fact store_sales ss_sales_price
Fact catalog_sales cs_sales_price
Fact web_sales ws_sales_price
Calculation · BI Algebra: conformed_measure Sales UNION ALL across channels · channel-aware
Calculation · BI Algebra: Growth YoY % Change = Growth(Sales, year) growth(x, period) = (xcurrent − xprior) / xprior templates: growth · ratio · rate · benchmark · time_compare +20.9% Books · Jan 1999 · store channel

05 / Discovery

Ask for a direction. Get options to explore.

Sometimes the right question is "show me what's interesting." Spotonix proposes four candidate Segmentations — each with its own BI Algebra — and lets you drill into any of them.

> Show me how the customer dimension is segmented with respect to sales.

Three customer-related dimensions surfaced from your Context Graph — all joined to store_sales, catalog_sales, web_sales. Each attribute below is an axis you can segment on; no values fetched yet.

Dimension

customer

identity & lifecycle

  • c_preferred_cust_flag 2 values flag_partition
  • c_birth_year date cohort_window
  • c_first_sales_date_sk date cohort_window

Dimension

customer_demographics

demographic profile

  • cd_gender 2 values cross_segment
  • cd_marital_status 5 values cross_segment
  • cd_education_status 6 values cross_segment
  • cd_credit_rating 4 values cross_segment

Dimension

customer_address

geographic · hierarchical

  • ca_state 50+ values dimension_drilldown
  • ca_city many dimension_drilldown
  • ca_zip many dimension_drilldown
  • Hierarchy country → state → city → zip

06 / Compound

Every verified Answer densifies your Context Graph.

Watch five real questions play through. The Segments and Calculations from each Answer persist — and the reuse rate climbs as the graph fills in. The 100th question is dramatically faster than the 1st because most of its building blocks already exist.

See it with your data.

From a 60-second sandbox to a 30-minute guided session with a founder.

Fastest · 60 seconds

Live Sandbox

A working Spotonix instance against a sample retail dataset. Ask any question. No login.

Try it now →

Most Relevant · 5 minutes

Upload Your PBIX

See how Spotonix interprets your own DAX models — dry-run, no warehouse credentials required.

Coming soon · request early access →

Deepest · 30 minutes

Guided Demo

30 minutes with a founder. Bring your questions, your data challenges, your PBIX files.

Book a slot →