About

One engineer, start to finish.

Darviq Systems is run by Ravi Shekhar Reddy, a cloud and platform architect in Bangalore working across cloud, AI and robotics.

Background

Ravi Shekhar Reddy

Before founding Darviq Systems I worked as a Technical Architect on large-scale AWS, OpenStack and Kubernetes platforms for enterprise and telecom customers: leading the move of production microservices from OpenStack to Amazon EKS, automating environments with Terraform and Ansible, and working onsite with international customers as their main technical contact.

I founded Darviq Systems to offer that same platform work to smaller teams and to build products of our own across cloud, AI and robotics: Darviq-Health for clinics, with AI drafting for doctors; Darviq-Observability, automated monitoring across Prometheus, Grafana, Loki, Zabbix, Datadog and Splunk; Darviq-Humanoid, a robotics simulation with a learned controller; and Darviq-Quantum, a quantum computer simulated from scratch. Find me on LinkedIn.

Ravi Shekhar Reddy
Experience

Large-scale production platforms.

Enterprise and telecom-scale platforms on AWS, OpenStack and Kubernetes, including onsite delivery with international customers.

IoT at scale

Cloud platforms for connected devices in very large numbers, built and deployed to run around the clock.

IoTAWS

Migrations to Kubernetes

Production microservices moved from OpenStack to Amazon EKS in phases, with tested cutovers and rollback plans.

OpenStackAmazon EKS

Automation

Environments described in Terraform and Ansible, so a new one takes minutes instead of days and every change is reviewed.

TerraformAnsible

Lower running costs

AWS and Kubernetes spend brought down by right-sizing clusters and removing waste, without giving up reliability.

AWSKubernetesCost
How I work

Run it, measure it, then write it down.

1Test against failure. Stop a service, pull a node, push past the limits, and watch what actually happens before calling anything done.
2Automate the repeatable. If it will be done twice, it goes in Terraform, Ansible or a pipeline.
3Say what isn't done. Every handover lists what's finished, what's left and what to watch.
4Leave the team able to run it. Documentation and a walkthrough are part of the job, not extras.

Let's talk about your platform.

A short call is the quickest way to find out whether I can help.