Nurudeen Kamilu
Senior Systems Engineer @ Visa | Kubestronaut | Championing Reliable Container Infrastructure
Warsaw, Poland
Actions
Nurudeen Kamilu is a highly respected Systems Engineer with years of hands-on experience architecting, securing, and scaling containerized infrastructures across the financial, aviation, health, and technology sectors. He is passionate about bridging Dev and Ops through GitOps, IaC, observability, and SRE practices, with a focus on AI-assisted operations, agentic workflows, and intelligent automation to improve reliability, responsiveness, and business outcomes.
Nurudeen has consistently delivered enterprise-grade platforms across on-premise infrastructure, AWS, GCP, and Azure, and is known for driving automation, resilience, and policy governance in production-grade clusters. A CNCF-recognized Kubestronaut, Nurudeen is also the only Black recipient of the Kubestronaut award in Poland, a testament to his technical excellence and community impact. Don’t be surprised—like Linux distros, Nurudeen has deep knowledge of OpenSource Kubernetes, EKS, GKE, Mirantis Kubernetes Engine (MKE), Oracle Kubernetes Engine (OKE), and OpenShift.
Beyond his professional responsibilities, Nurudeen is deeply committed to the tech community. He actively shares his knowledge through speaking engagements and mentorship programs, aiming to uplift and educate aspiring engineers. Nurudeen's passion for technology and education extends to his involvement with organizations like DevOps-as-a-Service (DaaS), where he contributes to fostering a culture of innovation and excellence within the DevOps community.
Whether he's building a secure cluster or mentoring aspiring engineers, Nurudeen brings energy, insight, and authenticity to every stage and conversation.
Links
Badges
Area of Expertise
Topics
The Truth About GPU Sharing in Multi-Tenant Kubernetes
We’ve all been there: you finally secure a high-end GPU for your cluster, only to realize that assigning a massive card to a single small pod is a huge waste of money. So you try to share it. But GPU sharing today requires a careful strategy because when it’s misconfigured, things quickly turn into chaos. A single rogue AI training job can crash an entire GPU node, taking down multiple production inference pods along with it.
This session is targeted to address exactly these challenges. We’ll break down how GPU sharing actually works and why it often doesn’t. We’ll take a deep dive into the illusion of resource isolation provided by GPU time-slicing and NVIDIA Multi-Process Service (MPS), which is fast but risky in multi-tenant environments. We’ll also discuss why vGPU frequently introduces more operational and management overhead than it solves.
Finally, we’ll bridge the gap between hardware capabilities and orchestration by exploring how Dynamic Resource Allocation (DRA) combined with Multi-Instance GPU (MIG) enables true hardware isolation. This approach allows GPUs to be dynamically partitioned into isolated slices based on real workload demand.
By the end of the session, attendees will leave with a clear blueprint for building a multi-tenant GPU cluster that delivers strong isolation for security and stability, while still achieving high utilization and predictable performance through dynamic partitioning.
Securing the Software Supply Chain with Kyverno, Cosign, and OPA
As the Kubernetes ecosystem matures, so do the threats targeting its software supply chain—from insecure container images to misconfigured workloads and unverified deployments. This talk dives into how platform and DevOps engineers can implement real-time, in-cluster controls to secure their Kubernetes environments beyond the CI pipeline.
We’ll walk through practical, production-ready implementations of:
1. Cosign for signing and verifying container images
2. Kyverno for applying policy-as-code and enforcing trusted sources
3. OPA (Open Policy Agent) for fine-grained admission controls and compliance
4. How to combine these tools to detect, prevent, and respond to supply chain risks at deploy time
The session includes live examples of rejecting unsigned images, enforcing namespace-specific rules, verifying image provenance, and layering policies for secure multi-team clusters. Attendees will leave with a clear blueprint for securing software delivery pipelines from build to cluster.
Not All Schedulers Are Created Equal: Why Your ML Jobs Deserve Better Than Default Kube-Scheduler
Kubernetes has become the de facto platform for deploying machine learning (ML) workloads, but its default scheduler was designed for long-running web services—not batch jobs, distributed training, or GPU-heavy tasks. As a result, teams running ML workloads often face poor cluster utilization, job starvation, and costly inefficiencies—all because the scheduler isn’t optimized for their needs.
In this talk, we’ll explore:
1. Why the default kube-scheduler falls short for ML workloads (e.g., no gang scheduling, poor spot instance handling).
2. How specialized schedulers (like Apache Yunikorn, Kube-batch, or Volcano) solve these problems with features like topology-aware placement, fairness queues, and elastic scaling.
3. Real-world case studies of teams that improved job completion times and reduced costs by switching schedulers.
4. A decision framework for choosing the right scheduler based on workload requirements.
Multi-Scheduler Kubernetes: A Deep Dive into Flexible Workload Orchestration
Kubernetes, the industry-leading container orchestration platform, has evolved to support a wide array of workloads. However, the default Kubernetes scheduler may not always be optimal for complex or highly specialized workloads. This session will delve into the power and flexibility of running multiple schedulers within a single Kubernetes cluster. By leveraging custom schedulers, operators can tailor the scheduling behavior for diverse workload types, improve performance, and address unique resource needs. We will explore the configuration, management, and integration of multiple schedulers, discussing strategies such as scheduler priorities, custom scheduling policies, and scheduler tuning to meet the demands of modern applications.
Nurudeen Kamilu
Senior Systems Engineer @ Visa | Kubestronaut | Championing Reliable Container Infrastructure
Warsaw, Poland
Links
Actions
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top