Snowflake Data Cloud Summit 2024 marked a significant milestone in the company's evolution towards becoming an enterprise AI powerhouse, featuring major partnership announcements, new AI-focused products, and a bold vision for the future of intelligent data applications. The summit was rich with thought-provoking presentations that audience members were eager to bring back to their companies to gain a competitive edge, generating tremendous excitement among Snowflake’s growing community of customers, partners, and developers.
Snowflake’s CEO and former head of AI, Sridhar Ramaswamy, emphasized the company’s vision to evolve its data cloud into a turnkey solution for data storage, data pipeline development, AI and ML model creation and maintenance, application development, serverless computing, and more. Snowflake highlighted its platform’s ability to simplify the complexities of sharing and utilizing massive amounts of data, in a low-code no-code fashion, while still offering the ability to develop highly customized features as needed. Users of all skill-level can spin up and execute data workloads and applications quickly within a unified platform without the need to move data around or create excessive security layers. A recurring theme throughout the summit was that while using a given feature, all other features should be seamlessly available no matter where a user is in the platform. The features are interconnected and work together seamlessly, making it feel like a single, cohesive product. As AI and data-driven applications become commonplace, Snowflake is pioneering the field and bringing a clear strategic vision to the table.
Many product announcements and keynotes at Summit focused on enabling AI and ML innovation, ultimately empowering companies with out of the box features in the Snowflake platform. This was illustrated as Jensen Huang, president and CEO of NVIDIA, joined Sridhar Ramaswamy remotely for the opening keynote. The two leaders discussed the future of AI and ML work, emphasizing their commitment to advancing Snowflake’s platform with NVIDIA’s software and hardware.
A key highlight was Cortex AI which is a code studio for creating and maintaining AI and ML models within a few clicks. To illustrate an example, from the Snowflake home screen a forecasting model can be created and connected to an existing time series data set within a few clicks. The result is a SnowSQL script complete with functional code, comments, and is ready for sequential execution. After execution, forecast results are ready in a view for consumption. It’s truly as simple as that – and since the result is in a commented script – adjustments and edits can easily be made to fit the use case at hand.
The initial version of Cortex AI is complete with three out of the box ML models – forecasting, anomaly detection, and classification – and the ability to create entirely custom scripts. Companies can create applications leveraging serverless infrastructure and rollout RAG based LLMs, ML models, chatbots, and industry specific AI solutions within days – not months.
Beyond easy model creation, Snowflake offers a range of major LLMs, including their in-house model Snowflake Arctic, Llama3, Mistral, Gemma, and more—all within one platform. This eliminates the need to move data around or create excessive security layers.
To further enable generative AI, Snowflake highlighted the abilities of SnowflakeCo-Pilot and Cortex Analyst. Co-Pilot uses natural language to facilitate data analysis, helping users answer questions about data and its structure. Its features include efficient query writing, optimization recommendation, troubleshooting, and assisting in data exploration. Cortex AI is geared towards business users to allow them to find answers quickly and analyze their data efficiently. It can be embedded in internal and external applications and enhance time to value for data-driven applications.
In addition, Snowflake emphasized their integration of Snowpark Container Services (SPCS) and Snowflake native applications. Container services are not new but have come a long way since their inception in early 2023. At its core, SPCS offers to run anything that can run in a docker container in Snowflake. Developers can build applications in any language, package, and deploy code in containers with any configurations of CPU/GPU and manage container images all inside the Snowflake environment.
The enhancement of the native application framework allows data application providers to create Snowflake-integrated applications for internal use, public use, or use by a specific partner or set of partners. Snowflake’s integration of Streamlit, a simple yet effective python framework for developing custom web applications, made way for democratized data applications. With features like Streamlit and container services, application providers can develop an application, package the application in a container image, and share with application consumers that can run the application it in their own environments with their own compute pools. Ultimately, this will bolster the Snowflake marketplace and will support the rise in data driven applications for both public and private use.
A compelling example of this was a RAG-based application developed with Streamlit, Mistral, and container services. The app was designed to scrape YouTube videos, store the video data, and allow users to prompt an LLM to answer questions about the video. The engineer leading the session illustrated how to run an open-source vector DB (Weaviate) entirely within a Snowflake container and have it interact with the Streamlit application to fetch relevant response about the prompt.
Snowflake unveiled Snowflake Notebooks, a coding environment that closely mirrors Jupyter Notebooks. Developers can write SQL, Python, and markdown language in a familiar IDE. Additionally, Snowflake announced the ability to convert Snowflake DataFrames to Pandas DataFrames. This provides the familiar, pandas-native experience that many developers know and love. By enabling Pandas code to be compiled and run with Snowflake compute ensures high performance, without the need of dedicated servers elsewhere. For those familiar with Pandas and Jupyter Notebooks, these enhancements are significant and will drive Snowpark’s adoption beyond the early adopter stage. Moreover, the introduction of Snowflake Dark Mode – a UI aesthetic feature – is a welcome addition that many users will appreciate. a UI aesthetic feature – is a welcome addition that many users will appreciate.
The links below offer detailed information on the major announcements made at the Snowflake Summit 2024. While this article covers some of the highlights, there is additional information on many announcements that are not discussed here.
AI & ML
Data Management & Integration
Developer Tools & Frameworks
Observability
Cleanrooms have been around for a long time, and Snowflake is revitalizing this concept, recognizing its significant potential in the Martech and Adtech space. In 2023, Snowflake acquired the clean room technology Samooha, which boasts an intuitive UI and focuses on reducing the complexity of creating and maintaining data cleanrooms. The technology is integrated into Snowflake’s UI, which allows users to create a cleanroom in under 20 minutes. Once created, the cleanroom is not siloed from Snowflake’s other feature. Users can still create row access policies, implement masking policies, run stored procedures, and treat their clean room as if it were another table or view in their data warehouse.
As the Snowflake Data Cloud Summit 2024 came and went, it's clear that the company is not just keeping pace with the rapidly evolving landscape of AI and data-driven applications but is actively shaping it. With groundbreaking product announcements, strategic partnerships like the ones with NVIDIA and AWS, and a vision that seamlessly integrates AI, ML, and advanced data management, Snowflake is poised to lead the industry into the future. The excitement and innovation showcased at the summit highlight a community ready to leverage these new tools and capabilities, driving forward a new era of intelligent data applications. Whether you’re a developer, data scientist, or business leader, the advancements unveiled at this year’s summit offer powerful new ways to harness data and unlock value.
Snowflake Data Cloud Summit 2024 marked a significant milestone in the company's evolution towards becoming an enterprise AI powerhouse, featuring major partnership announcements, new AI-focused products, and a bold vision for the future of intelligent data applications. The summit was rich with thought-provoking presentations that audience members were eager to bring back to their companies to gain a competitive edge, generating tremendous excitement among Snowflake’s growing community of customers, partners, and developers.
Snowflake’s CEO and former head of AI, Sridhar Ramaswamy, emphasized the company’s vision to evolve its data cloud into a turnkey solution for data storage, data pipeline development, AI and ML model creation and maintenance, application development, serverless computing, and more. Snowflake highlighted its platform’s ability to simplify the complexities of sharing and utilizing massive amounts of data, in a low-code no-code fashion, while still offering the ability to develop highly customized features as needed. Users of all skill-level can spin up and execute data workloads and applications quickly within a unified platform without the need to move data around or create excessive security layers. A recurring theme throughout the summit was that while using a given feature, all other features should be seamlessly available no matter where a user is in the platform. The features are interconnected and work together seamlessly, making it feel like a single, cohesive product. As AI and data-driven applications become commonplace, Snowflake is pioneering the field and bringing a clear strategic vision to the table.
Many product announcements and keynotes at Summit focused on enabling AI and ML innovation, ultimately empowering companies with out of the box features in the Snowflake platform. This was illustrated as Jensen Huang, president and CEO of NVIDIA, joined Sridhar Ramaswamy remotely for the opening keynote. The two leaders discussed the future of AI and ML work, emphasizing their commitment to advancing Snowflake’s platform with NVIDIA’s software and hardware.
A key highlight was Cortex AI which is a code studio for creating and maintaining AI and ML models within a few clicks. To illustrate an example, from the Snowflake home screen a forecasting model can be created and connected to an existing time series data set within a few clicks. The result is a SnowSQL script complete with functional code, comments, and is ready for sequential execution. After execution, forecast results are ready in a view for consumption. It’s truly as simple as that – and since the result is in a commented script – adjustments and edits can easily be made to fit the use case at hand.
The initial version of Cortex AI is complete with three out of the box ML models – forecasting, anomaly detection, and classification – and the ability to create entirely custom scripts. Companies can create applications leveraging serverless infrastructure and rollout RAG based LLMs, ML models, chatbots, and industry specific AI solutions within days – not months.
Beyond easy model creation, Snowflake offers a range of major LLMs, including their in-house model Snowflake Arctic, Llama3, Mistral, Gemma, and more—all within one platform. This eliminates the need to move data around or create excessive security layers.
To further enable generative AI, Snowflake highlighted the abilities of SnowflakeCo-Pilot and Cortex Analyst. Co-Pilot uses natural language to facilitate data analysis, helping users answer questions about data and its structure. Its features include efficient query writing, optimization recommendation, troubleshooting, and assisting in data exploration. Cortex AI is geared towards business users to allow them to find answers quickly and analyze their data efficiently. It can be embedded in internal and external applications and enhance time to value for data-driven applications.
In addition, Snowflake emphasized their integration of Snowpark Container Services (SPCS) and Snowflake native applications. Container services are not new but have come a long way since their inception in early 2023. At its core, SPCS offers to run anything that can run in a docker container in Snowflake. Developers can build applications in any language, package, and deploy code in containers with any configurations of CPU/GPU and manage container images all inside the Snowflake environment.
The enhancement of the native application framework allows data application providers to create Snowflake-integrated applications for internal use, public use, or use by a specific partner or set of partners. Snowflake’s integration of Streamlit, a simple yet effective python framework for developing custom web applications, made way for democratized data applications. With features like Streamlit and container services, application providers can develop an application, package the application in a container image, and share with application consumers that can run the application it in their own environments with their own compute pools. Ultimately, this will bolster the Snowflake marketplace and will support the rise in data driven applications for both public and private use.
A compelling example of this was a RAG-based application developed with Streamlit, Mistral, and container services. The app was designed to scrape YouTube videos, store the video data, and allow users to prompt an LLM to answer questions about the video. The engineer leading the session illustrated how to run an open-source vector DB (Weaviate) entirely within a Snowflake container and have it interact with the Streamlit application to fetch relevant response about the prompt.
Snowflake unveiled Snowflake Notebooks, a coding environment that closely mirrors Jupyter Notebooks. Developers can write SQL, Python, and markdown language in a familiar IDE. Additionally, Snowflake announced the ability to convert Snowflake DataFrames to Pandas DataFrames. This provides the familiar, pandas-native experience that many developers know and love. By enabling Pandas code to be compiled and run with Snowflake compute ensures high performance, without the need of dedicated servers elsewhere. For those familiar with Pandas and Jupyter Notebooks, these enhancements are significant and will drive Snowpark’s adoption beyond the early adopter stage. Moreover, the introduction of Snowflake Dark Mode – a UI aesthetic feature – is a welcome addition that many users will appreciate. a UI aesthetic feature – is a welcome addition that many users will appreciate.
The links below offer detailed information on the major announcements made at the Snowflake Summit 2024. While this article covers some of the highlights, there is additional information on many announcements that are not discussed here.
AI & ML
Data Management & Integration
Developer Tools & Frameworks
Observability
Cleanrooms have been around for a long time, and Snowflake is revitalizing this concept, recognizing its significant potential in the Martech and Adtech space. In 2023, Snowflake acquired the clean room technology Samooha, which boasts an intuitive UI and focuses on reducing the complexity of creating and maintaining data cleanrooms. The technology is integrated into Snowflake’s UI, which allows users to create a cleanroom in under 20 minutes. Once created, the cleanroom is not siloed from Snowflake’s other feature. Users can still create row access policies, implement masking policies, run stored procedures, and treat their clean room as if it were another table or view in their data warehouse.
As the Snowflake Data Cloud Summit 2024 came and went, it's clear that the company is not just keeping pace with the rapidly evolving landscape of AI and data-driven applications but is actively shaping it. With groundbreaking product announcements, strategic partnerships like the ones with NVIDIA and AWS, and a vision that seamlessly integrates AI, ML, and advanced data management, Snowflake is poised to lead the industry into the future. The excitement and innovation showcased at the summit highlight a community ready to leverage these new tools and capabilities, driving forward a new era of intelligent data applications. Whether you’re a developer, data scientist, or business leader, the advancements unveiled at this year’s summit offer powerful new ways to harness data and unlock value.
Lorem ipsum dolor sit amet consectetur. Sit sit vel eu sed proin. Arcu bibendum in mauris urna erat magna volutpat eget. Lectus commodo consectetur egestas quis ut lobortis nunc.
.png)
Lorem ipsum dolor sit amet consectetur. Sit sit vel eu sed proin. Arcu bibendum in mauris urna erat magna volutpat eget. Lectus commodo consectetur egestas quis ut lobortis nunc.
.png)
%20(1).jpg)
.jpeg)
