Kasun Talwatta
Kasun “Kas” Talwatta
@day0ops
New Zealand

Twenty years in software leaves a particular kind of mark. I started as afull-stack developer, building across industries that had little in common: fleet and job tracking, health IT, enterprise systems. They all shared the same fundamental challenge though, which was distributed data and the infrastructure underneath it. The parts that fascinated me most were never the features. They were the seams between systems and what happened when those seams failed.

That path eventually led to where I am now. I work as a Principal Architectat a cloud-native company, working closely with customers to untangle complex problems and turn them into solutions that actually hold up. A lot of the role is about meeting people where they are: understanding what they're trying to build, what's in the way, and helping them find a path through it. Platform strategy and distributed systems design are still at the core, but the work is as much about people and problems as it is about technology.

Open source is less a preference and more a conviction. The best ideas endure because they're visible, testable, and improvable by anyone. I've contributed where I could and learned far more from reading other people's code than from any course or textbook.

Platform & Cloud
KubernetesService MeshIstioCiliumPlatform EngineeringAWSMulti-cluster
Observability
PrometheusOpenTelemetryDistributed TracingeBPFGrafana
DevOps & GitOps
Argo CDFluxTerraformCI/CDSRE
AI Infrastructure
LLMsAgentic SystemsMLOpsGPU InfrastructureRAGVector Databases
Architecture
Solution ArchitectureDistributed SystemsAPI DesignSystem DesignCloud Native
Where it started
The languages from the early years. Still close to the metal, still something I genuinely enjoy.
JavaC++CPythonShellPerlRuby

Outside of terminals and architecture diagrams, I tinker withelectronics and robotics. Hardware debugging has a different kind of satisfaction. There is no stack trace, but the smoke tells you everything.

The name day0ops has three readings:

day zero opsFoundational infrastructure work. The stuff that has to be right before anything else can run.
zero-day mindsetSecurity-aware thinking baked in from the start, not bolted on after the breach.
day0-oopsBecause things break. That's where the learning is. The orange dot is always on fire. That's not a bug.
QR code for x.com/day0ops
X / Twitter
x.com/day0ops
QR code for email contact
Email
~/career  ·  git log --graph --stat
❯ git log --graph --stat --author="day0ops" --since="20 years ago"
c4a91f3Architect · Solution Architecture~3 years ago → now
Designing systems before they exist, and living with the consequences
From hands-on operations to shaping the platforms others run on. Solution architecture at this level means you're accountable for decisions that surface as 3am incidents two years later; that's why the ops background matters.
The AI infrastructure shift landed right here: designing for a world where GPU scheduling, model serving, and agent orchestration are first-class infrastructure concerns.
+Platform strategy across cloud-native and AI-native stacks
+Solution design for distributed, multi-cluster, multi-tenant systems
+Technical leadership, making trade-offs explicit
solution architectureplatform engineeringservice meshistioAI systemscloud strategydistributed systems
7f3d2b8DevOps · SRE~10 years ago
Kubernetes, service mesh, and the 3am education
The move from writing software to running it was the most humbling shift in 20 years. Kubernetes cluster operations at scale, service mesh (Istio, Linkerd, the sharp edges of both), and the SRE discipline of treating ops as an engineering problem.
Production taught me more about software design than any design review. Every incident became a post-mortem, every post-mortem became a blog post.
+Kubernetes cluster operations; multi-cluster, production at scale
+Service mesh; Istio, Linkerd, mTLS, traffic management, observability
+SRE practices; error budgets, runbooks, toil reduction, on-call
kubernetesistioservice meshSREobservabilityCI/CDgitopsincident response
1a0e5c2Java Developer · Software Architect~20 years ago
Java, enterprise systems, and learning to design software that outlives you
Started where most engineers start, writing code and figuring out why it breaks. Java development through the enterprise era: J2EE, Spring, distributed systems before Kubernetes was a thing.
The move into software architecture meant owning decisions, not just implementations, and learning that the cost of a bad design compounds with every engineer who has to work around it.
+Java development; Spring, J2EE, enterprise apps
+Software architecture; system design, API design, technical debt
+CS fundamentals; algorithms, data structures, distributed systems
javaspringsoftware architecturesystem designdistributed systemsAPI designopen source