Vanna AI - Empowering Data Teams through Chat-GPT
We all know what Chat-GPT is and what it has done in the past year, it has been everywhere in the tech world. Today, I am going to share how Data Teams, Data Users can get boost their day to day workflows by adopting the Chat-GPT based
Text to SQL tool called Vanna.ai.
Vanna AI provides an abstraction on top of Chat-GPT that allows users to easily integrate any Database or Data Warehouse and train models leveraging the Chat-GPT powers and getting results back in the format that end users love.
How Vanna AI works
A quick glimpse of Vanna, no more manual SQL is needed.
To learn more about Vanna, its features and capabilities, check it on website.
In this article, I will share the four use cases that I think Vanna AI can add value and you should consider.
I have not looked into any other alternative, there are likely to be many other similar services now.
Accelerating Data Exploration
Vanna AI can help teams to speed up their data analysis and exploration journey by a big factor, teams who spend hours in writing complex SQL queries now can easily and quickly get the answers and focus on the business problems they are solving, whether it be just an answer to a question, or understanding the data. This can really change the game the way we work with Data today.
Enabling Non-Technical Teams
There has been scenarios when non technical teams like Finance, Executives, Legal and many other teams that may ask such questions that are easy to answer but require someone with SQL expertise, this can enable them to ask it directly to a computer instead of a human by interacting with Vanna AI in natural language.
vanna.ask("What are the top 10 customers by sales?")
Typically, we solve these cases through dashboards, Vanna AI also provides graphs and charts of the returned data.
Elevating SQL Skills
This is a great way to learn, improve and polish your SQL skills, since Vanna returns the SQL script along with the actual data (It also returns suggestions like data visualization and follow-ups questions, check video above), it becomes easy to see and learn how natural language like English is translated into a data language to talk to a data warehouse. From beginners to experts, everyone can benefit from Vanna.
An example of query taken from Vanna docs.
SELECT c.c_name as customer_name,
sum(l.l_extendedprice * (1 - l.l_discount)) as total_sales
FROM snowflake_sample_data.tpch_sf1.lineitem l join snowflake_sample_data.tpch_sf1.orders o
ON l.l_orderkey = o.o_orderkey join snowflake_sample_data.tpch_sf1.customer c
ON o.o_custkey = c.c_custkey
GROUP BY customer_name
ORDER BY total_sales desc limit 10;
Adopting as its Cool
Chat-GPT is cool, everyone is using, and everyone is talking about it, why not adopt and add it to tech stack, this would definitely add value to company especially when hiring folks.
Furthermore, if you want to look more cool, you can utilize the powers and train on your custom data make it more specific for your workflows. This is supported by Vanna, its very easy to train.
vanna.train(sql="SELECT * FROM my-table WHERE name = 'John Doe'")
On top of Vanna, Engineers can also build their layers and integrate with custom platform to provide additional data services.
Let me know what other scenarios you think should be considered in the comments.
If you are looking to learn more about Vanna or want to get started, check out their open source library, they also have a hosted version.
For more thorough guide on setting up, read this article Ditch SQL — Let Vanna Handle It!