Snowflake Data Cloud Summit 2024: Key Announcements and Major Developments Unveiled

Every year, Snowflake hosts the Data Cloud Summit to showcase advancements in its platform, helping teams leverage their data assets effectively. In 2024, the focus was on a crucial aspect: empowering teams to build robust AI applications using their data.

Led by CEO Sridhar Ramaswamy, Snowflake announced several innovative AI capabilities and partnerships, alongside the launch of long-anticipated features aimed at enhancing the platform's data foundation. Highlights include the general availability of Iceberg Tables and an internal marketplace.

Major Announcements from the 2024 Snowflake Data Cloud Summit:

1. Iceberg Tables and Polaris Catalog

Snowflake made Iceberg Tables generally available, offering full storage interoperability for enterprises. These tables function like native Snowflake tables but store metadata in Apache Iceberg format within customer-supplied storage. This innovation enhances Snowflake’s usability, performance, governance, and collaboration for externally stored Iceberg data. Moreover, the launch of Polaris, an open data catalog, facilitates the indexing and organization of data in open table formats and supports interoperability with various query engines, including Apache Flink, Apache Spark, Dremio, Python, and Trino.

2. Cortex AI and Snowflake ML Enhancements

Cortex AI, Snowflake’s fully managed service for building large language model (LLM) applications, received significant upgrades, including an AI & ML Studio with a no-code interface for easy application development. New features include Analyst and Search offerings for LLM chatbot development and retrieval augmented generation (RAG) applications. To ensure safety, Cortex Guard will prevent chatbots from generating harmful content, alongside enhanced MLOps capabilities in Snowflake ML.

3. Snowflake Horizon's New Marketplace

Snowflake introduced Horizon, which includes features for compliance, security, privacy, and data governance, and has enhanced it with a new internal marketplace. This marketplace allows users to curate and publish data products tailored for their teams while controlling access to sensitive content. It supports sharing AI models, Iceberg Tables, and Dynamic Tables, along with AI-powered object descriptions to improve data discovery.

4. Snowflake Trail for Observability

Snowflake launched Trail, an observability tool that enhances visibility into data quality, pipelines, and applications. Trail provides telemetry signals for Snowpark and Snowpark Container Services, allowing users to diagnose and debug errors effortlessly without manual data transfers. Utilizing OpenTelemetry standards, it integrates seamlessly with observability platforms like Datadog, Grafana, Metaplane, PagerDuty, and Slack.

5. Enhanced Data Residency

Snowflake announced an expansion into regulated and sovereign markets, including a new EU-only data boundary to enhance data residency and help customers comply with regulations. A dedicated environment for Department of Defense customers will offer heightened security and networking integration with Boundary Cloud Access Point (BCAP).

6. Partnership with Nvidia

Snowflake formed a partnership with Nvidia to adopt Nvidia AI Enterprise software and incorporate NeMo Retriever microservices into Cortex AI. This collaboration allows organizations to connect custom models with business data for more accurate responses. The open LLM Arctic supports Nvidia TensorRT-LLM software, optimizing performance and accessibility through NVIDIA NIM inference microservices.

The Snowflake Data Cloud Summit is set to run from June 3 to June 6, 2024.