JFrog is releasing a brand new end-to-end resolution for growing and deploying enterprise AI functions that brings collectively growth groups, information scientists, and machine studying engineers right into a single platform.
JFrog ML supplies a holistic view of all the AI software program provide chain, from software program packages to LLMs, in order that corporations can guarantee their AI functions are secured in the identical approach their conventional software program is.
It supplies safety scanning for AI fashions, whether or not they had been created in-house or are from a third-party.
Different key options embody offering a single system of document, reproducible artifacts for all fashions created within the platform, simplified mannequin growth and deployment processes, and dataset administration and have retailer assist.
“Because the demand for AI-powered functions continues to develop quickly, so do the considerations across the capacity to manage and handle this new area on all fronts – from MLOps to ML safety. In reality, our personal crew of safety researchers had been the primary to seek out and assist remediate new, zero-day malicious ML fashions in Hugging Face,” stated Alon Lev, VP and GM of MLOps at JFrog. “JFrog ML combines superior, easy and hassle-free consumer expertise for bringing fashions to manufacturing, mixed with the extent of belief and provenance enterprises anticipate from JFrog, permitting clients to speed up their AI initiatives with confidence.”
JFrog companions with Hugging Face, NVIDIA
The corporate additionally introduced new partnerships with Hugging Face and NVIDIA. As a part of the Hugging Face partnership, fashions in Hugging Face will now be scanned by JFrog and can show a “JFrog Licensed” checkmark if they’re deemed protected.
In response to JFrog, its mannequin scanner goes deeper than different scanners, which has allowed it to eradicate 96% of false positives produced by different scanners on Hugging Face fashions and to establish malicious fashions not flagged by different scanners.
“Because the main collaboration platform for AI fashions, we’re delighted to deepen our partnership with JFrog to implement high-quality scanning capabilities for our AI/ML fashions and ship higher peace of thoughts for builders trying to create the subsequent era of AI-powered functions,” stated Julien Chaumond, CTO of Hugging Face.
NVIDIA NIM microservices can also be now built-in into JFrog’s platform, which additional permits safe mannequin supply by providing clients enterprise-grade safety, elevated visibility, and governance controls.
“Efficiency and safety are essential for profitable enterprise AI deployments,” stated Pat Lee, vice chairman of Enterprise Strategic Partnerships at NVIDIA. “With NVIDIA NIM built-in immediately into the JFrog Platform, builders can speed up AI adoption with a unified, end-to-end resolution for constructing, deploying, and managing manufacturing AI brokers at scale.”