Enkrypt AI
Kubernetes and GPU infrastructure for an AI platform, with reliable model-serving workloads on AWS.
KubernetesAWSGPUHelmPrometheus
view engagement →system.identity
Harsh Yadav
Site Reliability Engineer / DevOps Engineer
Building reliable, scalable, cloud-native infrastructure.
KubernetesAWSTerraformObservabilityAutomation
session
$ whoami
harsh@production
role: associate-sre @ cloudraft
experience: 1 year
focus: cloudnative, aws, azure
status: open-to-workprojects.current
Client engagements
Hearzap
Scaling and hardening the microservice platform behind an audio/social app, focused on reliability and observability.
KubernetesAWSHelmTerraformGrafana
view engagement →SkySwitch
Cloud-native delivery and platform work for a SaaS product, automating deployments and stabilizing production.
KubernetesAWSCI/CDArgoCDDocker
view engagement →blog.latest
Recent writing
Cloud Native Software Distribution for Enterprise AI
Discover how SaaS companies can build enterprise-ready software that runs securely in customer VPCs, private clouds, and air-gapped Kubernetes clusters.
·15 min read
Debugging Post-Upgrade Timeouts: The Case of Stale Endpoints in EKS
After an EKS 1.32 to 1.33 upgrade, internal services started timing out while every pod looked healthy. The culprit was stale Endpoints pointing at dead pod IPs. Here is how we found it and fixed it.
·3 min read
Feature Flagging
Improve reliability in platform engineering with advanced feature flagging. Enable safe deployments, progressive delivery, and real-time experimentation to streamline modern software delivery.
·17 min read