Exclusive: Voltron Data Acquires Claypot to Enable Real-Time AI Capabilities with Modular Data Systems

Voltron Data Acquires Claypot to Enhance Real-Time Analytics

San Francisco-based Voltron Data, a startup specializing in modular and composable data analytics solutions, has officially announced its acquisition of real-time AI platform Claypot. The financial details of the transaction remain undisclosed.

This strategic acquisition will integrate Claypot's talented team into Voltron, enabling the expansion of real-time analytics capabilities. Enterprises will now have the tools to unlock substantial value from their most current data, enhancing applications such as fraud detection and personalized marketing.

Collaboration in Real-Time Streaming Data

Josh Patterson, co-founder and CEO of Voltron Data, stated, “Our two teams have collaborated for eight months to create the first streaming data backend. We are excited to welcome Claypot’s founding members to Voltron Data, where we will continue to innovate in streaming and unify real-time with batch data processing. Our potential for growth together is immense.”

Leveraging Open Source Technologies

Founded in 2022 with $110 million in funding, Voltron Data draws inspiration from the animated series Voltron, where distinct components unite to form a powerful entity. By harnessing leading open-source technologies such as Apache Arrow, Apache Parquet, and Ibis, Voltron facilitates improved data access and analytics.

The core mission is to empower teams to select the best components at each layer of the data stack, allowing the construction of modular systems that are adaptable to evolving organizational needs.

Recently, Voltron launched its Theseus distributed query engine, designed to accelerate CPU-intensive preprocessing tasks like ETL, feature engineering, and transformation using a cluster of Nvidia GPUs and other hardware accelerators. This advancement allows for seamless integration of data preprocessing and AI/ML operations, streamlining data analytics and AI pipelines on a unified infrastructure.

Claypot's Dual Processing Capabilities

Claypot, founded by former Nvidia and Netflix engineers Chip Huyen and Zhenzhong Xu, enables enterprises to utilize the freshest data for accurate predictions and rapid insights. The platform adeptly combines streaming and batch data processing, catering to various enterprise demands by balancing latency, cost, and accuracy.

“Claypot works with both streaming and batch data, allowing you to select the best-performing data for each use case. If a delay of one hour is acceptable, we can switch to batch processing. However, if immediate responses are needed, we’ll transition to streaming,” the company highlights on its website.

With Claypot's technology and expertise, Voltron can now enhance its modular offerings, previously centered on local and batch data, to provide comprehensive analytics from streaming data while upholding open standards. This integration will empower enterprises with real-time AI, feature engineering, and MLOps capabilities driven by Theseus and open-source products.

Patterson emphasized the goal of providing system builders greater flexibility with the Claypot AI team's integration, stating, “We will extend our vision of composability and hardware acceleration further into real-time data and machine learning operations.”

Aiming to Eliminate Technology Silos

While the exact timeline for Claypot's integration is still to be determined, Voltron is optimistic about dismantling technology silos in data analytics and AI, with additional initiatives in development.

“We have foundational work ahead. Two significant use cases are preprocessing structured and semi-structured data for ML/AI models, and managing real-time data streams that require low latency. Current workloads are often too large and sensitive for Spark and Presto. Although we do not have further acquisitions planned right now, we continuously seek top talent aligned with our vision to accelerate our growth,” Patterson added.

Since its inception, Voltron Data has partnered with “numerous” clients across key sectors, including federal agencies, to enhance query performance on petabyte-scale datasets. In November 2023, Voltron partnered with HPE to incorporate Theseus into HPE Ezmeral unified analytics software.

“Our customers are reaping the rewards of our accelerated solutions, tackling larger and more critical data challenges with newfound ease,” Patterson concluded.