How you can now create SQL with Natural Language

Google launches Duet AI in BigQuery

How Google uses AI to ease Data Analysis via SQL

Google has launched Duet AI for BigQuery, which is an AI-powered collaborator in Google Cloud. It should help Data Scientists, Analysts and especially business users to complete, generate, and explain SQL queries[1].

Similar to how Microsoft has already integrated AI such as ChatGPT for Azure Synapse & Co. and how you can use them to create source code, SQL queries and data pipelines by entering normal language, Google is now following suit and offering something similar for the BigQuery Data Warehouse.

Let’s take a look at this new feature and how it exactly works. In order to use it, you have to compose a new query in the query editor, enter one of the following natural language prompts and then press Enter. Here is a small example for that instance. Use the official documentation linked down below to dive deeper into this topic[2]:

# Using `bigquery-public-data.austin_bikeshare.bikeshare_trips`, calculate the
# average trip length by subscriber type.

BigQuery then will create a SQl-query, based on your natural language input via text:

SELECT
subscriber_type,
AVG(duration_minutes) AS average_trip_length
FROM
`bigquery-public-data.austin_bikeshare.bikeshare_trips`
GROUP BY
subscriber_type

So this is a pretty nice new feature, especially for users who need to perform SQL Analysis via BigQuery but are maybe not so experienced in SQL (yet). Also, Google wants to probably keep up with Microsoft in terms of services and products. What you have to keep in mind is that this feature is currently in preview mode. Hence, it could be not available for everybody for now.

Other also very useful updates which also came in lately can be found down below:

Sources and Further Readings

[1] Google, BigQuery release notes (2023)

[2] Google, Write queries with Duet AI assistance (2023)