JFrog acquires Qwak AI to streamline the development and production of AI models

”

JFrog will continue to expand its solutions to bring advanced MLOps capabilities to enterprises, enabling them to build, deploy, manage, and monitor AI workflows including generative AI, large language models (LLMs), and conventional machine learning models on a unified platform.

JFrog (NASDAQ: FROG), a streaming software company and creator of the JFrog software supply chain platform, today announced that it has entered into a definitive agreement to acquire Qwak AI, creator of the AI and MLOps platform.

With this acquisition, JFrog aims to provide a unified, scalable solution for DevOps, security, and MLOps stakeholders. This industry-leading, state-of-the-art MLOps capability is designed to free data scientists and developers from infrastructure issues and accelerate the creation and delivery of AI-driven applications. JFrog is the single system of record for all software packages (binaries), including the models stored in Artifactory. Enhancing its machine learning (ML) model capabilities will further enable users to streamline the entire process of model development to deployment.

Shlomi Ben Haim, CEO and co-founder of JFrog, said, "Next-generation software supply chain platforms need to extend and natively include MLOps solutions to better serve development organizations. We are excited to integrate Qwak's MLOps solution into our platform to empower our customers' AI journeys." With JFrog Artifactory as the preferred model registry and JFrog Xray for scanning and protecting ML models, Qwak's solution continues to focus on improving user efficiency and providing a unified platform experience for DevOps, DevSecOps, MLOps, and MLSecOps. We look forward to working with the team at Qwak to make it even greater! ”

As part of the JFrog platform, Qwak technology will bring an intuitive and simple user experience to the model production process, ensuring the high level of trust and traceability that enterprises expect when deploying AI-driven applications. The combination of Qwak's advanced model training and service capabilities enables the management of independent and complex model lifecycles, combined with JFrog's model storage management and security scanning services.

The acquisition follows the successful integration of JFrog and Qwak earlier this year, which is based on JFrog's model as a package approach. This holistic solution is designed to eliminate the need for separate tools, simplify compliance, and enable full traceability through a single solution.

Alon Lev, CEO and Co-Founder of Qwak, said, "We are thrilled to be joining the JFrog family and helping our customers accelerate their AI initiatives. We founded Qwak with a vision to change the way software development teams and data scientists work together to bring AI assets into production. By leveraging the power of the JFrog Software Supply Chain Platform to deliver secure software components at scale, we're creating a new experience that aims to pave the way for unified digital delivery teams to bring responsible, secure models to their applications in a simpler and more predictable way. ”

As enterprises begin to deliver AI-powered applications, ML models as the driving force behind AI, enabling their rapid time-to-market and secure flow are key factors behind modern MLOps initiatives. According to Gartner, MLOps plays a critical role in AI operations, with 75% of companies expected to move from AI pilots to operations by the end of 2024.

"Data scientists and ML engineers are currently using tools that are mostly disconnected from standard DevOps and security processes within the company, delaying release times and eroding trust," said Gal Marder, executive vice president of strategy at JFrog. A unified system of record across development, security, machine learning, and operations will address this pain point for digital teams and businesses. ”

Today's market demands a unified platform experience across the software supply chain to accelerate the development process and address key AI drivers such as ML models and metadata. Like other software components, ML models must be stored, built, tracked, versioned, signed, secured, and efficiently delivered across systems to deliver AI applications at scale. Leveraging DevOps practices in a unified solution can meet these market expectations.

The acquisition of Qwak will expand the JFrog solution with the following capabilities:

• Unified platform for DevSecOps and MLSecOps: Provides a holistic ML software supply chain from traditional models to large language models and generative AI

• Deliver model services to production quickly and directly: Optimize AI initiatives by further streamlining model development, deployment, and servicing

• Model training and monitoring: Implement applications with OOTB dataset management and feature storage support

• Manage the model as a package: Enable customers to version, manage, and secure the model in the same way as other software packages by using DevSecOps best practices

• Guaranteed security: Proof of origin and security of AI are naturally ensured in the development workflow

• Extract models from controlled, secure, and real-world sources: Integrate ML models with other building blocks such as containers and Python packages

• Model traceability: Easy recall, retraining, and redeployment of production models in the event of a problem

JFrog's MLOps development program

As part of the acquisition and integration process, JFrog plans to rapidly grow its MLOps-centric team by bringing Qwak's talent to the JFrog team. JFrog will also accelerate technology integration and bring Qwak technology to the JFrog platform across JFrog DevOps and security products. JFrog and Qwak will work closely with customers to ensure business continuity and a smooth transition to future jointly developed and supported products.

The MLOps ecosystem integrates with partners

Earlier this year, JFrog announced an integration with MLflow, developed by AWS Sagemaker and DataBricks. As part of the company's quest for versatility, JFrog will continue to offer integrations with other leading MLOps ecosystem partners, giving developers and ML engineers greater freedom of choice and avoiding the risk of vendor lock-in.

About JFrog:

JFrog Ltd. (NASDAQ: FROG) is on a mission to create a frictionless world of software delivery from developer to device. With a "streaming software" philosophy, the JFrog Software Supply Chain Platform is a unified system of record that enables organizations to build, manage, and distribute software quickly and securely, ensuring that software is available, traceable, and tamper-proof. Integrated security capabilities also help identify, defend against, and remediate threats and vulnerabilities. JFrog's hybrid, general-purpose, multi-cloud platform is available as a self-hosted and SaaS service across multiple major cloud service providers. Millions of users and more than 7,200 customers around the world, including most Fortune 100 companies, rely on JFrog solutions for secure digital transformation. Use it and you'll know! For more information, please visit jfrogchina.com or follow us on WeChat: JFrog.