At swampUP 2025, Yuval Fernbach, vice president and CTO of MLOps at JFrog, and Adel El Hallak, senior director of product for NVIDIA AI Enterprise, discuss how JFrog is integrating machine learning operations (MLOps) into its platform strategy. Fernbach, who joined JFrog through its acquisition of Qwak, outlined how JFrog is extending its mission beyond artifact management to embrace AI-driven software supply chain automation. The goal, he noted, is to enable developers to build, deploy, and govern AI models with the same level of trust and traceability that JFrog has long brought to binary and container management.

They discussed the rapid convergence of MLOps and DevOps, a shift that is redefining how teams manage the entire software and model delivery lifecycle. As organizations accelerate their adoption of generative AI and large language models (LLMs), the challenge is no longer just building or training models—it’s governing and managing them at scale. AI introduces new forms of technical debt, compliance requirements, and operational risk that traditional software management frameworks were never designed to handle. Every model carries dependencies, training data, and inference patterns that must be tracked, versioned, and secured across environments. Without unified governance, enterprises risk exposing sensitive data, violating regulations, or deploying unverified models that behave unpredictably in production.

The energy at swampUP reflects an industry in transition—where every team, from developers to platform engineers, is rethinking their toolchains for the AI era. As Fernbach and others shared, JFrog’s latest innovations are designed to bridge that gap, ensuring that software—and now models—move from experimentation to production with speed, security, and confidence.