I had the opportunity to attend the Snowflake AI Data Cloud Summit 2024, and it was an incredible experience. I was excited to be part of an event hosting around 20,000 passionate data and AI enthusiasts from around the globe. Over four days, more than 500 sessions covered every conceivable aspect of Data and AI across various industries.
Snowflake CEO Sridhar Ramaswamy’s tagline, “Era of Enterprise AI is here,” perfectly captured the summit’s spirit. According to him, AI-driven innovation is set to dominate the next decade, and synergy with strategic partners, like the one between Snowflake and NVIDIA, is crucial.
The event showcased Snowflake’s roadmap and innovations in AI, highlighting how enterprises are already benefiting from these advancements. I particularly enjoyed the theme of ease of use, enabling various personas to harness AI. In this blog, I discuss three key Snowflake developments that stood out to me, along with my learnings and key takeaways.
1 Strengthening the Data Foundation
I was impressed by Snowflake’s focus on enabling data access on your terms, leveraging AI to work seamlessly with unstructured data, and supporting open formats outside of Snowflake. The emphasis on data governance, discovery of data, apps, and models was notable. Snowflake’s capabilities promise optimal total cost of ownership (TCO) and continued performance optimization, with a 27% improvement in query duration being particularly noteworthy.
Here are some of the most exciting features and upcoming enhancements announced at the summit that are focused on strengthening the data foundation:
General Availability (GA) Features
- Dynamic Tables: A declarative approach to transformations, joining, and aggregating across multiple source objects, automating incremental refresh with low latency.
- Iceberg Tables: An open table format that brings Apache Iceberg to Snowflake, allowing data storage outside Snowflake with great performance and a single pane of glass for data discovery.
- Cost Management Interface: A centralized interface for managing costs in Snowflake.
- Snowflake Horizon: A tool to govern and discover all objects with AI-powered capabilities and a simplified UI.
- Snowflake Data Clean Rooms: Secure, collaborative environments for data sharing.
- Universal Search: Enhanced search capabilities.
- Advanced Analytics: Time Series (ASOF JOIN): Improved analytics for time series data.
Upcoming Features
- Document AI: A fully managed workflow using Snowflake Arctic-TILT for seamless unstructured data handling with a no-code UI.
- Polaris Catalog: An open-source catalog for Apache Iceberg with cross-engine read and write interoperability.
- Data Quality Monitoring: : A native UI for system metrics and custom metrics for automated data quality measurement.
- Lineage Interface: : Interactive lineage visualization with manual tag propagation workflow.
- Trust Center: A UI for discovering security risks and recommendations.
2 Accelerating Enterprise AI
Working closely with open-source models, I am amazed by how accuracy is improving while costs are decreasing. Despite this progress, reducing the cost of experimentation and failing fast with AI remains a challenge. Snowflake has made significant strides in developing, deploying, and operationalizing ML features and models. With secure access to top-tier LLMs and AI services for building chatbots and native integrations into SQL, Python, and REST APIs within the Snowflake ecosystem, we can expect rapid AI implementation to drive business value.
Here are some of the most exciting features and upcoming enhancements announced at the summit, all focused on accelerating enterprise AI.
General Availability (GA) Features
- Snowflake Model Registry: Manage and share AI/ML models natively within Snowflake.
- Cortex Guard: Safeguard applications with filtering to prevent undesirable outputs, built with Llama Guard.
- Cross-Region Inference: Access inferences across regions.
Upcoming Features
- Snowflake Notebooks: An easy-to-use Notebooks interface with integrations like Snowpark ML, Streamlit, Cortex, and Iceberg.
- Snowpark Pandas API: An extension adding distributed pandas at scale.
- Cortex Search: : A managed text search solution (semantic + keywords).
- Cortex Analyst: Enables business users to interact with structured data using natural language.
3 Building & Distributing Applications
Having spent over a decade in application development before transitioning to the Data and AI world, I’ve encountered numerous challenges moving data to applications for analytical and transactional needs. With the exponential growth in data volumes, it makes sense to bring application development to the data. Snowflake’s focus on making Data and AI accessible to businesses and the ability to easily build, distribute, and monetize full-stack apps in the AI data cloud were highlights. Native application performance monitoring was the cherry on top.
Here are some exciting Snowflake developments in the realm of building and distributing applications.
General Availability (GA) Features
- Snowpark Container Services: Bundle apps in a container and bring them to Snowflake (supports both CPU and GPU compute instances).
- Data Pipeline Observability: A native UI for monitoring key metrics throughout the data pipeline.
- Serverless Tasks: Define tasks to run on serverless infrastructure from Snowflake.
Upcoming Features
- Native Apps Integration with Snowpark Container Services: Build Snowflake-native apps using containers and languages of your choice.
- Snowflake Python API: An interface to interact with all Snowflake resources.
- GIT Integration: Leveraging full CICD power in the Snowflake ecosystem.
- Snowflake CLI: Commands to manage the lifecycle of apps running on Snowflake.
- Database Change Management: Declarative management of all Snowflake objects with code across environments.
In Summary, Snowflake AI Data Cloud Summit 2024 showcased numerous groundbreaking developments that are set to transform the way enterprises manage and utilize data. From strengthening data foundations to accelerating AI and optimizing application development, Snowflake’s advancements offer powerful tools to drive business success. With a wealth of these capabilities now available, implementing AI solutions has never been more accessible or cost-effective, allowing businesses to experience its transformative potential with minimal resource investment.
Infocepts: Enabling Enterprises to Become Truly Data-Driven
At Infocepts, we empower enterprises to harness the full potential of modern data platforms like Snowflake. With our deep expertise in data and AI, we guide organizations to leverage cutting-edge capabilities to build a robust data foundation, accelerate AI initiatives, streamline application development, and more. Speak with an expert today to learn how Infocepts can support your modernization journey and enable your organization to fully unlock the power of your data.
Prashant is a Data and AI enthusiast with over 17 years of experience in designing and building applications and systems that power enterprise Data & Analytics Platforms. He is proficient in multiple programming languages and has co-created numerous enterprise-grade products in collaboration with major Cloud (Infra & Data) vendors.
Read Full BioRecent Blogs
From Vision to Reality: Building a Successful DEI Program
August 29, 2024
Data Storytelling for Data Management: A 7-Part Blog Series by Scott Taylor
August 29, 2024
AI in Sales: Building a Compelling Business Case for Enhanced Results
August 21, 2024
Top Data and AI Breakthroughs (and Controversies) of 2024
August 21, 2024
Infocepts
