7 Layers: VMware Focused on Open Source Interaction, ML Security

Chip Childers has been at VMware for six months, during which time he has focused on steering the vendor’s use and interaction of open source software platforms and refining its approach to supporting those efforts. These are significant areas of interest for VMware and for Childers’ role as VP and chief open source officer at the vendor, which is increasingly interacting with the open source community. “For projects that VMware hosts, we want to make sure that we're very welcoming to participants that want to collaborate with us, but also how do we engage in a really authentic way with open source communities that are out there, maybe at a foundation or owned by another vendor or just even a collective of individuals,” Childers said. “So we spent a lot of time on making sure that VMware has a very thoughtful community strategy and how it engages with these projects.” Those comments build on what Childers said were his initial areas of focus when he took on the role last year. At that time, Childers repeatedly mentioned a need for VMware to be “intentional” with its interactions across the open source community. This work also includes VMware exploring other open source opportunities that it can help foster and in turn help power new VMware services. “We're exploring project communities, we're trying to make a very positive impact in those communities and, frankly, get to the point where those communities are generating software that we think is going to be useful for VMware, either internally or as part of our product strategy,” Childers added. This includes a specific focus on operationalizing machine learning (ML). Childers mentioned projects like Kubeflow, which supports ML stacks running on Kubernetes. Kubeflow is currently under the guise of Google, but is working toward inclusion in the Cloud Native Computing Foundation (CNCF). Childers also touched on the growing need for VMware to work on securing open source ML platforms. He described attack models that can add noise to an image in a way that can trick an ML model into thinking “a banana is an apple.” To counter this, VMware is looking at different projects, with Childers citing the Adversarial Robustness Toolbox project that is currently on GitHub as a potential way to test various threats or attack vectors. “As we look at it we see how important these operational attributes of machine learning are going to be to our customers and, frankly, to the industry writ large,” he added. Listen to more of Childers’ insight into VMware’s areas of focus in the latest episode of the SDxCentral 7 Layers podcast. Learn more about your ad choices. Visit megaphone.fm/adchoices