Peter Salanki
Chief Technology Officer at CoreWeave
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolPeter Salanki is a seasoned technology executive and CTO with 12 years of experience leading cloud infrastructure and engineering teams, currently steering CoreWeave’s technical strategy. He progressed through senior engineering roles at CoreWeave from Director to VP before becoming CTO, and earlier held architecture and leadership positions across networking and security firms where he designed carrier-grade networks and data center systems. Hands-on with cloud-native and virtualization projects, he contributes to prominent open-source efforts such as KServe (serverless ML inference on Kubernetes) and KubeVirt, improving production robustness, autoscaling, and VM resource efficiency. Based in New York, he blends strategic leadership with deep systems-level expertise—comfortable both defining platform roadmaps and diving into asyncio thread-safety, autoscaling annotations, or Virtio-FS optimizations. Notably, his background spans low-level network engineering to modern MLOps, giving him a rare end-to-end view of cloud performance and operational reliability.
12 years of coding experience
21 years of employment as a software developer
Enskilda Gymnasiet
Swedish, English
Stackoverflow
11reputation
373reached
0answers
1question
Github Skills (24)
kubernetes10
docker10
mlops10
go10
tensorflow10
virtualization10
kubernetes-pods10
istio10
knative10
develop9
libvirt9
python9
asyncio9
backend9
automations8
Programming languages (20)
Github contributions (5)
kserve/kserve
May 2020 - Sep 2020
Standardized Serverless ML Inference Platform on Kubernetes
Role in this project:
MLOps Engineer
MLOps EngineerContributions:2 reviews, 11 commits, 10 PRs in 3 months
Contributions summary:Peter primarily focused on improving the deployment and operational aspects of the KServe platform. They modified the code to manage container concurrency, disabled retries in Istio, and added liveness probes to Tensorflow predictors. Further contributions involved updating KNative annotations related to autoscaling, refactoring KFServer to support asynchronous HTTP requests and fixing related thread safety issues in the asyncio implementation, demonstrating a focus on enhancing the system's robustness and performance in a production environment.
xgboostsklearnserverlessclient-goknative
kubevirt/kubevirt
Oct 2020 - Sep 2021
Kubernetes Virtualization API and runtime in order to define and manage virtual machines.
Role in this project:Back-end Developer
Back-end DeveloperContributions:5 reviews, 9 commits, 6 PRs in 11 months
Contributions summary:Peter contributed to the Kubevirt project by addressing issues related to virtual machine initialization and resource management. They focused on optimizing the `virt-launcher` Pod, disabling service links and Virtio-FS metadata cache to reduce memory usage and improve performance. Furthermore, they removed an unnecessary `FSGroup` setting and fixed an issue with DHCP server, and they also worked on persistent volume claim configuration to ensure correct behavior with Virtio-FS. The user's work demonstrates their understanding of Kubernetes, virtualization technologies, and system optimization within the Kubevirt framework.
apivmsmachineslibvirtruntime
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.