Data analysis is changing fast. Tools like Snowflake make it easier to handle large amounts of data. But what if you could query Snowflake using natural language? Enter DataLine.
DataLine is an AI-driven open source and privacy-first platform for data exploration through chat. This makes data analysis more accessible, especially for those who aren’t data experts. It’s easy to install and stores all your data locally making it great for self-hosting and using in businesses where data security is crucial. Today we’ll be plugging it into a Snowflake database.
That’s enough of an intro, I know you’re excited to get your hands dirty. Let’s get to it.
Connecting with Snowflake
Today we’ll be working with the Snowflake samples, as that requires no setup. You can create a new Snowflake account if you don’t have one already.
Jump into your Admin -> Account settings and let’s get our host name.
Here you can see the host name is ‘jcaiijn-qr76787’. Other than that, we just need our username and password (Snowflake user password), and we’re all set!
For a username of ‘bob’ and password of ‘thebuilder’, the DSN (database connection string) here would be:
snowflake://bob:thebuilder@jcaiijn-qr76787/SNOWFLAKE_SAMPLE_DATA/TPCH_SF1
Plugging in to DataLine
Now all you need to do is add your database connection string from above in to DataLine.
Playtime
We’re already here? Whew.. that was fast.
You can just type your questions in plain english and it’ll figure out what needs to be done. Here I’ll explore this sample DB with DataLine’s help. Let’s start with asking for some example questions so we get an idea about what’s in there:
Woah cool! That would’ve taken a data expert at least 10 minutes to get an idea of what the data looks like and go through all the tables. DataLine translated your question into several Snowflake queries, retrieved the tables and analyzed them. You see the answer in seconds but there was a lot of work behind the scenes!
There’s more?
DataLine doesn’t stop at answering questions. It also helps you visualize the data. Let’s see what the average order quantity per customer is, and ask for a chart of the distribution per customer for the top 10 customers.
Ok seriously though, how long would it have taken a data analyst to lookup the schemas, write this query, export the data, and build such a pretty chart? That’s what this is all about. Saving tons of time and tons of money and making this more accessible to non-technical folks. It was always about the question, not about how we get to an answer. We were just drowning in the details.
AI really is transforming data analysis, in a good way. With tools like DataLine, the focus shifts from writing complex queries to asking the right questions. Now:
- Non-experts can analyze data without learning SQL or other query languages
- We get insights faster by skipping the query-writing process.
- (Eventually) AI will minimize errors in data interpretation compared to humans. Not there yet but not that far off.
Aand that’s a wrap! DataLine and Snowflake together make AI data analysis more intuitive and powerful. We’re always looking for help building this awesome open-source product, and we have big plans for its future. If you’re interested, you know where to find us.
