At Snow Day, held on the 18th, Snowflake announced technology that will enable developers, data engineers, and data scientists to build applications directly on the data cloud.

This announcement, which will change the landscape of app development, will enable more users to leverage data, increase productivity, and develop applications, pipelines, and machine learning (ML) models on Snowflake's single platform.

■ Streamlit acquisition supports Python-based app development in the data cloud

Last March, Snowflake acquired Streamlit, a data app development platform, enabling countless data scientists and developers to easily build Python apps with the open-source framework. With this, Snowflake plans to combine Streamlet with the Snowflake Data Cloud to enable developers to build data and ML models into secure, interactive applications.

This allows developers to easily build and deploy web apps, offering the convenience and flexibility of Streamlet and the controlled data access, scalability, and security of Snowflake. Developers can build, deploy, and operate Python apps using their own data on the secure Snowflake platform. This allows them to seamlessly share data with business teams and enhance the utilization of ML.

“Streamlit serves as a bidirectional engine for JetBlue’s data science and ML, enabling us to build apps with our business unit counterparts and collaborate more efficiently,” said Sai Ravuru, Head of Data Science and Analytics at JetBlue Airways, which is using Streamlit. “Snowflake’s acquisition of Streamlit will allow JetBlue to leverage both data and ML insights on the Snowflake platform.”

Based on this, it was reported that the flight planning team will increase predictability through a powerful app to reduce damage from unexpected situations such as flight cancellations, and provide the best customer experience with safer and more convenient customized customer service.

Expanding the Python Ecosystem with 'Snowpark for Python'

Python, ranked as the third most popular programming language among data scientists, is expanding its reach to all users through its expanded business ecosystem. Snowflake has released 'Snowpark for Python', providing access to Python and a rich ecosystem of open source libraries.

The new SnowPark will provide developers with a flexible architecture that natively supports their preferred programming language among Java, Scala, SQL, and Python.

Among these, "Snowpark for Python" enables multiple teams to collaborate and build apps on a single platform within a highly secure Python sandbox, offering high scalability, flexibility, security, and consistent governance when building apps. This allows developers to address data security and compliance issues that previously hindered productization.

Additionally, Snowflake plans to enable Python developers to run large-scale ML training and memory-intensive operations directly on the Snowflake platform through a "Snowpark Dedicated Warehouse," available as a public preview through Amazon Web Services (AWS). The company also announced plans to release a private preview of "Python Worksheets," enabling developers to develop applications, data pipelines, and ML models within Snowflake.

Several partners, including Anaconda and dbt Labs, have been instrumental in accelerating the adoption of Snowflake for Python and enabling developers to easily build apps. Anaconda's Snowflake for Python eliminates the need for manual processing and package dependency management, giving Snowflake users seamless access to Anaconda's open-source Python libraries. Additionally, dbt Labs' Snowpark for Python combines the power of SQL and Python for modern analytics, helping customers bridge the gap between data analysts and data scientists.

“Snowpark for Python gives financial engineers, data scientists, and business analysts the integrated data science ecosystem they need to build and deliver customized analytics on diverse data sets,” said William Woo, head of quantitative analytics at Northern Trust. “They can collaborate with their business units on the same data, discover and visualize new insights, and more effectively support clients’ investment decisions based on dynamic facts and data.”

■ Streamlining the streaming pipeline to enhance developer automation and visibility.

Snowflake revolutionizes the way data pipelines are built, simplifying work with streaming data on a single platform and eliminating silos. To achieve this, Snowflake leverages core software development principles while leveraging essential capabilities to eliminate complexity.

Users can now work with their data faster and be more productive with schema inference, available in private preview, and easily run pipelines with serverless jobs, available to all users. Along with this, we are releasing improved tools to strengthen developers' ability to build apps on the data cloud.

“As Snowflake continues to build apps, we’re providing developers with the data access and tools they need to accelerate innovation—all on a single, secure platform,” said Torsten Grabs, Head of Product Management at Snowflake. “Snowflake’s enhanced offering empowers developers to confidently build powerful applications, pipelines, and models, while removing complexity so they can leverage the data cloud across the enterprise and drive business value.”

“Snowflake’s Python release has opened up new opportunities and use cases for the NerdWallet team to build and deploy secure, compliant pipelines on Snowflake, enabling us to more efficiently provide our customers with the tools they need for all their financial workflows,” said Sathish Balakrishnan, Director of Data Engineering at NerdWallet. “Snowflake’s continued investment in Python gives NerdWallet the flexibility to code in their preferred programming language and accelerate the pace of innovation for our end users.”