Mick McGuinness
Mick McGuinness DBmarlin Co-founder & Product Manager

DBmarlin AI Co-pilot saves Enerj time tuning their databases

DBmarlin AI Co-pilot saves Enerj time tuning their databases

DBmarlin has recently introduced a new AI Co-pilot feature, which eliminates the need for experts to spend hours manually analysing and fine-tuning complex queries. Enerj, a long-term DBmarlin user across various versions of PostgreSQL and Microsoft SQL Server, quickly adopted this new functionality and experienced immediate benefits.

Enerj leads in providing solutions and services for electronic document preservation and E-invoicing. With offices in Italy, France, and England, they serve over 6000 clients in sectors such as Distribution, Transport and Logistics, Services and Finance, Automotive, Agro-Food, Fashion, and more.

How Enerj used DBmarlin AI to save time

Gabriele D’Onufrio, an expert DBA with years of experience, currently works at Enerj and is based in Veneto, Italy. The following scenario demonstrates how Gabriele effectively used DBmarlin AI Co-Pilot to save time, receiving a PostgreSQL index recommendation without requiring any time-consuming manual analysis.

Enerj identified a slow, high-cost SQL statement using DBmarlin. The SQL text and execution plan in DBmarlin are shown below.

enerj-copilot-screen1.png

Co-pilot recommendations

Using DBmarlin AI Co-pilot they were able to get the following recommendations:

enerj-copilot-screen2.png

enerj-copilot-screen3.png

After applying the suggested index the new execution plan now looks like this in DBmarlin with the cost of the most expensive step dropping from 69,185 to just 3.

enerj-copilot-screen4.png

Comparing the SQL statistics in DBmarlin before and after the index creation, we can see that the Physical reads dropped from 314K to 24K.

enerj-copilot-screen5.png

enerj-copilot-screen6.png

About DBmarlin AI Co-pilot

As well as index suggestions like in the example above, DBmarlin AI Co-pilot can also make many other types of recommendations:

  • Create new index - including the DDL for the CREATE INDEX statement
  • Rewrite SQL query
  • Update table and index statistics
  • Use partitioned tables
  • Use materialised views
  • Tune a database parameter
  • Scale up hardware

co-pilot.gif

If you would like to find out more about DBmarlin AI Co-pilot or try out DBmarlin head to https://www.dbmarlin.com/dbmarlin-ai-co-pilot where you can find videos and links to a freemium version to get you started.

dbmarlin-community-join