Snowflake Nabs Streamlit for Data Science Apps

Cloud data warehousing giant Snowflake showed it’s serious about Python and data science this week when it announced that it plans to spend $800 million to buy Streamlit, a provider of Python-based tools for rapidly developing interactive data applications on the Web.

Co-founded in San Francisco in 2018 by Adrien Treuille, Amanda Kelly, and Thiago Teixeira, Streamlit develops an open source framework of the same name that allows data scientists and machine learning engineers to create and deploy data applications. The software is compatible with other Python-based frameworks, such as NumPy, Pandas, Matplotlib, and Scikit-learn, and uses React to render screens on the front-end.

Since debuting about four years ago, Streamlit has collected a devoted following in the data science and machine learning departments at organizations like Stitch Fix, Uber, Yelp, and Google X. Streamlit’s most telling characteristic seems to be speed. Users appear to like how quickly they can go from the scripting stage in a Juypter notebook into a full-blown Web app, replete with all the requisite HTML, CSS, and JavaScript.

An experienced data scientist, Treuille was burdened by the time and effort it took to generate a functional app. “Every nontrivial machine learning project is eventually stitched together with bug-ridden and unmaintainable internal tools,” he wrote in the first Streamlit blog post. “So my old Google X friend, Thiago Teixeira, and I began thinking about the following question: What if we could make building tools as easy as writing Python scripts? We wanted machine learning engineers to be able to create beautiful apps without needing a tools team.”

That is essentially what Streamlit has done. The company has raised $62 million over two rounds before engaging with Snowflake in talks about selling the company. Those talks led to Wednesday’s announcement that Snowflake plans to buy the company for $800 million. The deal is subject to regulatory approval and customary closing conditions.

Streamlit will be instrumental in building a community of data applications for the Snowlake Data Cloud, according to Benoit Dageville, a Snowflake co-founder and its president of products.

“We just realized this is exactly the right step for Streamlit,” said Adrien Treuille, Streamlit co-founder and CEO, on the acquisition by Snowflake 

“The iPhone is a platform to run applications,” he said in a video interview with Treuille. “Snowflake is going to be a platform to run data applications. And of course these data applications are going to be built not by us, but by all the community. Everyone will build these that applications. We’ll put these data applications in our marketplace and [share] that application with the rest of the Snowflake ecosystem and our customers.”

In a way, it’s remarkable that the deal even occurred, according to Treuille, who called Streamlit a “labor of love” that was never meant to be a business.

“It was a side project. It was something that we thought was cool,” he told Dageville in the video. “And even now, as it’s grown to be used all over the world, we feel that fundamentally, both we and our users are tool builders we revel in the ability to make magic happen just by typing into a text editor.”

Snowflake has ruffled some feathers in the data world by being a closed environment that restricts what tools its users can use with the data they have stored in its proprietary data warehouse. A year ago, Dageville and his co-founders cautioned against falling into the trap posed by open standards, open formats, and open source.

But to its credit, Snowflake has been opening itself up to the outside world over the past year. It added support for Python a couple of months ago giving data scientists and engineers the ability to work in their favored language. It is also embracing Apache Iceberg, an open data format used with Parquet to provide traceability and governance for data stored in lakes.

While it’s not exactly clear yet how Snowflake will position its Streamlit assets, the acquisition of a well-regarded and popular open source member of the Python data community would seem to strengthen its open bona fides.

There’s no indication openness had much to do with Streamlit’s decision (but $800 million reportedly was one of them). For Treuille, the decision to partner with Snowflake seemed to come down to the company’s reputation for building good products.

“I just felt really that complementary sense of these are tool builders the moment we talked to you guys,” he says in the video. “These are people who love the exact same internals that we love. These are people who think that the same awesome apps are awesome.

“And I also got a very strong sense of Snowflake’s appreciation for the value of Streamlit technology and Streamlit’s community,” Treuille continues. “And so over the course of many conversations, it basically became obvious that there was an amazing possibility if we joined forces. And I really feel like every day that we look at this, it becomes more clear that we can do amazing work both on the open source side and for Snowflake customers. And for the Streamlit community, bringing this technology to the world.”

In other news, Snowflake also announced its fourth quarter and fiscal year 2022 results. The company reported $359.6 million in the fourth quarter, representing 102% increase from the fourth quarter of 2021. For the year, it reported $1.14 billion in revenue, a 106% increase from the fiscal year 2021.

Wall Street did not like the news, as its stock, which is traded on the NYSE under the ticker symbol SNOW, was down 16%. The stock, which is trading around $210 per share Friday, has lost almost half its value from its high in November 2021, when it peaked at nearly $402 per share.

Related Items:

Snowflake Adds Python Support with Winter Release

All Eyes on Snowflake and Databricks in 2022

Snowflake, AWS Warm Up to Apache Iceberg

R and Python: The Data Science Dynamic Duo