Databricks SQL

Lakehouse answers for the whole business, from forecasts to fraud.

Try Dot with Databricks SQLRead the docs

From sales tables to model scores, Dot writes the SQL and shows it.

Threadanalytics

David Mensah9:41 AM

@Dot How far off was our demand forecast last month, by category?

1 reply

DotAPP9:42 AM

8 steps · 45s · Access Online

9% off overall. Outdoor gear missed by 31%.

Source: Databricks SQL’s sales.demand_forecast.

Chart

Threadanalytics

Sarah Chen9:41 AM

@Dot Which production lines lost the most to unplanned downtime this quarter?

1 reply

DotAPP9:42 AM

10 steps · 56s · Access Online

Line 4, mostly from one recurring press fault.

Source: Databricks SQL’s plant.downtime_events.

Here is the data.

LineDowntimeUnits lost
Line 462 h18,000
Line 214 h4,100
Line 111 h3,200

Threadanalytics

Ryan Ortiz9:41 AM

@Dot Of the transactions our fraud model flagged last week, how many turned out to be real fraud?

1 reply

DotAPP9:42 AM

4 steps · 20s · Access Online

412 of 2,310, an 18% hit rate, down from 24% the week before.

Source: Databricks SQL’s risk.flagged_transactions.

Chart

Connect Databricks SQL

  • Service principal or personal token
  • Unity Catalog or Hive metastore
  • Runs on your SQL warehouse
  • Can learn from query history

More in Databases & spreadsheets

Connect your data.

Start for Free