Introducing the Industry’s First AI Coprocessor for the Data Cloud

We are thrilled to be making two exciting announcements. Today we have launched RelationalAI’s preview availability, as well as being named a Snowflake Snowpark Container Services launch partner. As an AI coprocessor, RelationalAI extends your data cloud to support graph analytics, reasoning, optimization, and other composite AI workloads. Starting today, you can access RelationalAI directly from Snowflake. Snowpark Container Services (private preview) enables you to execute RelationalAI’s engine entirely within your Snowflake account with the same ease of use, scalability, and unified governance of the Snowflake Data Cloud.

AI has advanced at a staggering pace over the last year. The breakthroughs in large language models (LLMs) have created an urgency for every company to adopt AI or risk being left behind. In the April 2023 Gartner survey, “Executive Pulse: AI Investment Gets a Boost From ChatGPT Hype,” 45% of business executives cited ChatGPT as a significant factor behind their own adoption of AI.

But up until now, implementing AI has required putting together a complex set of point solutions, each offering a piece of the puzzle and requiring its own tools, infrastructure, and skill sets. This makes it incredibly difficult for you to apply and operationalize AI across your business. Getting the most from the latest AI technology requires a new approach that brings together your data and business logic and provides an environment for an ensemble of AI techniques to come together all in one place, in a secure and governed way.

At RelationalAI, we believe that one place is the data cloud. The combination of data, AI, and knowledge graphs within your data cloud allows you to more easily adopt AI for faster and better decision making. This is why organizations are already integrating LLMs in their data clouds, and why we are introducing RelationalAI as the first AI coprocessor for data clouds. Using RelationalAI, you are able to integrate knowledge graphs and use multiple AI techniques to derive knowledge, enrich data, automate decisions, and drive consistent use of business concepts.

What is the AI Coprocessor?

Just as a graphics processing unit (GPU) helps a central processing unit (CPU) in a computer by enabling specialized workloads like graphics, gaming, and machine learning the RelationalAI service acts as a coprocessor for the data cloud, enabling workloads including graph analytics, business rules, optimization, and other composite AI workloads.

At the core of RelationalAI is a relational knowledge graph system, developed based on years of proven relational database technology and some of the latest database systems research. Knowledge graphs are models of real world concepts and the relationships between them. Our innovation is in developing a new approach to knowledge graph systems that adheres to the proven relational paradigm, brings business logic together with data, and is cloud-native, making the AI coprocessor compatible with your data cloud. This enables you to execute multiple composite AI workloads from within your data cloud, eliminating the need to manage data movement or copies of data.

If you are developing intelligent solutions without RelationalAI, you have to replicate data outside of your data cloud and use specialized technology like navigational graph databases, rules engines, and solvers. These point technologies reside outside the data cloud’s governance framework and are typically neither cloud native nor relational. With RelationalAI, you can build intelligent solutions entirely within your data cloud, eliminating redundancies and reducing complexity, costs, and time to value.

RelationalAI’s Capabilities

RelationalAI brings together the best of relational technology and knowledge graphs to create an AI coprocessor that expands the capabilities of your data cloud. As a cloud-native service, storage and compute are separated, giving you the ability to clone data instantly, keep workloads isolated, and elastically scale based on your needs.

Working with data is made easy by treating graphs and relational data as atomic, irreducible relations - think of these as data building blocks. This provides you the flexibility to project data as tables, graphs, JSON, tensors, or matrices, and to compose relations that meet the needs of any application. Using the latest research in join algorithms, RelationalAI makes it possible to execute complex joins and traverse deeply connected graphs without sacrificing performance.

Built-in algorithms make it easy to utilize a variety of composite AI capabilities to drive decisions from data. Graph analytics libraries offer you support for common graph analytics tasks including centrality, community detection, similarity, and path analysis. Rules and business logic can be installed to enable simple and complex reasoning. Prescriptive analytics is also made possible with the ability to express and solve optimization problems through both open source and commercial solvers.

These capabilities are made available through SDKs for popular programming languages like Python and Java, as well as through SQL so that you do not have to learn new languages or tools to get started.

Using RelationalAI with the Snowflake Data Cloud

With the introduction of the AI coprocessor, RelationalAI’s capabilities are now available for use with the Snowflake Data Cloud and with the latest language models. Through this combination, we’re helping customers build on their Snowflake investments and easily extend into new workloads.

RelationalAI's algorithms in Snowflake

Using RelationalAI with Snowflake is simple and straightforward. First, identify the tables and views you’d like to materialize into your knowledge graph. Next, materialize that data as a knowledge graph - this is done through Snowflake SQL statements with RelationalAI automatically maintaining the latest materialization of your knowledge graph. With the data transformed into a graph structure, you can execute graph analytic algorithms directly from Snowflake to enrich existing data with insights from your knowledge graph.

As we work to bring data, AI, and knowledge graphs together within data clouds, we are thrilled to be a launch partner for Snowflake’s new Snowpark Container Services. We believe this will pave the way for customers to quickly bring the latest AI capabilities into a single environment with a common governance, security, and data management model.

“RelationalAI is a strong partner for knowledge graphs and other composite AI workloads. Our partnership with RelationalAI enables customers to drive more value with advanced composite AI capabilities like graph analytics, rules-based reasoning and optimization, all within the Data Cloud. We are delighted that RelationalAI is enabling our customers to do more within their Snowflake environments.”
Jeff Hollan
Director, Product Management, Snowflake

Led by Our Customers

We are incredibly fortunate to be working with companies who are among the leaders in their respective industries. These organizations have guided our direction, and we’ve built the AI coprocessor in direct collaboration with them to address real-world business needs. See what organizations are achieving using RelationalAI in production for business-critical workloads today.

“Our business depends on extremely complex business logic with interdependent rules and ever-changing regulatory requirements. The ability for RelationalAI to compress business rules and scale processing is stunning. Using their relational knowledge graph system, we slashed legacy code by 90% and reduced processing times from over a month to several hours. We’re thrilled with the operational efficiency and our domain experts are now empowered to follow the logic and quickly pinpoint necessary changes.”
Tax Technology Leader
EY Financial Services Office
“Thanks to RelationalAI, we’ve created a breakthrough in building intelligent data applications that understand our customer interactions and behaviors. Acting as an AI coprocessor, RelationalAI enables us to enhance our semantic models and perform sophisticated analysis like graph analytics to understand latent patterns in our data. We’re already seeing an impact in areas like reducing fraud, which has translated into significant business value in the first phase of our engagement. We are excited for the additional impact we expect to see as we leverage RelationalAI in additional use cases.”
Mark Austin
VP of Data Science and AI at AT&T

Get Started today

The future of AI in the data cloud is here, and we can't wait to embark on this exciting journey with you. Join us as we unlock the full potential of your data cloud and pave the way for innovative solutions and accelerated growth.

To learn more about RelationalAI or to get started, please visit us here.