+
C
A 7-act story about one enterprise modernization

From a rejected $18M GSI quoteto a lit-up Snowflake wall, in 15 months.

Watch Maya Alfaro — a principal data engineer with a decade of Teradata BTEQ on her shoulders — hand Cursor her legacy ELT portfolio. Her team keeps the keyboard. Jordan Park still gates every merge. Credits start flowing on asset #1, not asset #911.

Click through seven chapters. Every one has an interactive beat. Total runtime is ~8 minutes — and it's the demo a Snowflake AE plays on their next modernization call.

Act 01

The Wall

Q2 revenue: 14 hours stale

Act 02

The Quote

$18,000,000 · 4 years · rejected

Act 03

Diagnosis

63,180 LOC · indexed in 4 minutes

Act 04

First Asset

daily_revenue_rollup.bteq → fct_daily_revenue

Act 05

Proof & Review

Cortex semantic diff · 0 row delta · Jordan approves

Act 06

Scale

911 assets · 15 months · $16M pulled-forward credits

Act 07

The Morning After

Maya sleeps through the night. Samira closes the expansion.

~$16M pulled-forward credits

33 months earlier than the GSI baseline — quota retirement now, not in 2030.

Win-rate lift vs Databricks + BigQuery

Speed-of-migration is where Snowflake loses. This demo flips the axis in 8 minutes.

Cortex + Snowpark pull-through

Modernized code lights up the native features that drive expansion — automatically.

Row-level equivalence + Cortex semantic diff + human-approved merges throughout.