DevSecOps and MLOps combined: JFrog acquires ML platform Qwak

JFrog, provider of the software supply chain platform of the same name, has completed the acquisition of Qwak AI. Founded around four years ago, Qwak specializes in supporting data scientists in building machine learning models to enable flexible, scalable and secure deployment of ML applications.

The step from MLOps to MLSecOps

With the integration of Qwak into its own DevSecOps platform announced at the beginning of the year, JFrog aims to close the gap between MLOps and DevSecOps. By combining Qwak's ML platform with the various JFrog tools, the entire process from the development of a prototype to the productive operation of ML models can be accelerated and fully secured, explains JFrog CTO Yoav Landman as part of the takeover announcement. Companies and their IT organizations will benefit from better collaboration between data scientists, developers and security teams when introducing complex ML models and GenAI applications.

According to CTO and Qwak co-founder Yuval Fernbach, JFrog and Qwak complement each other perfectly in several respects. The integration creates a common platform for MLOps and DevSecOps, because of which standardized pipelines for software development and ML model management can be set up. In addition, JFrog's comprehensive range of security tools helps to use vulnerability detection and risk minimization for ML models and all other software components. This opens up the next step from MLOps to MLSecOps and includes security aspects such as analyzing dependencies in real time with JFrog Xray, automatically blocking non-compliant components and enforcing compliance guidelines (curating models).

Open for further cooperation in terms of MLOps

With the complete acquisition of Qwak, JFrog is continuing its "Model as a Package" strategy. As well as the technology, Qwak employees will be integrated into JFrog as part of an MLOps-centric team. While neither party has disclosed further details of the acquisition, JFrog leadership assures that the company remains open to further integrations with MLOps technologies. The collaboration with AWS SageMaker and Databricks (MLflow), also announced this year, would continue with unchanged commitment. Further details can be found in JFrog's official statement and in a blog post by Yuval Fernbach.

(map)