INITIALIZING DEVYANSHU.KAUSHIK...

0%

SYSTEM READY

Devyanshu Kaushik

Senior DevOps
Engineer

Cloud Infrastructure · Kubernetes · CI/CD · Python

I design, automate and operate scalable cloud infrastructure, production Kubernetes environments and reliable deployment systems.

100K+

Engineering production systems handling 100K+ requests/minute.

Establish connection Explore the infrastructure
ALL SYSTEMS OPERATIONAL deploy/api · idle
REQ/MIN100,412 CPU42% PODS9 ◌ PORTFOLIO SIMULATION
scroll

Production, in numbers

0

REQUESTS / MINUTE

Production Kubernetes with load balancing, run independently at Nitro Commerce.

0

PRODUCTION SERVERS

Linux, across GCP and AWS.

0

YEARS EXP.

Backend, DevOps and AI.

0

INFRASTRUCTURE ENVIRONMENTS

GCP · AWS · Onsite

Engineer behind the infrastructure

DevOps Engineer with 4+ years of experience designing, deploying and managing cloud-native infrastructure, Kubernetes environments, CI/CD pipelines, monitoring systems and distributed databases.

Most of that time has gone into keeping production boring: systems that stay up under 100,000+ requests a minute, deployments nobody has to watch, and logs and dashboards that answer questions before anyone asks. Python is the glue, and lately the agentic AI systems at RaySuite AI have become the newest thing to keep running.

SYSTEM PROFILEv2026
NAME
Devyanshu Kaushik
ROLE
Senior DevOps Engineer
EXPERIENCE
4+ Years
LOCATION
Gurugram, India
FOCUS
Cloud / Kubernetes / Automation
CURRENT
RaySuite AI (Jun 2025 –)
EDUCATION
B.Tech CSE, K.R. Mangalam University (2022)

Infrastructure I build

Click any component to see what sits behind it.

Cloud topology

KUBERNETES // PRODUCTION

Cluster
CPU42%
MEMORY61%
PODS42
NODES3
STATUSHEALTHY

simulated portfolio values, not live telemetry

Data layer

PostgreSQL

Relational workloads

MongoDB

Document store for platform services

Cassandra

Distributed, high-write data

Redis

Caching layer

MariaDB

Relational workloads

Production systems

How changes ship, and how I know they are healthy. GitHub Actions pipelines that reduced deployment time and made releases zero-touch, plus centralised monitoring and logging.

Deployment pipeline

  1. CODEcommit
  2. GitHubpush · PR
  3. GitHub Actionsworkflow
  4. Buildcompile
  5. Dockerimage
  6. Testchecks
  7. Deployrelease
  8. Kubernetesrollout
  9. Productionlive

PIPELINE IDLE

Scroll here to trigger a release.

Observability

Grafana and Loki, with log retention on Google Cloud Storage. Panels below are simulated for illustration.

REQUESTS100K+/min
LATENCYstable
ERROR RATElow
SERVICEShealthy
request rategrafana-style panel · simulated
log streamloki-style panel · simulated
  • Grafana
  • Loki
  • Centralized Logging
  • Monitoring
  • Log Retention
  • GCS Storage

Technology constellation

Hover a technology to pull it forward.Swipe through the stack.

DEVOPS

Kubernetes

Production clusters serving 100K+ requests/minute.

  • Platform & ops
  • Cloud & data
  • Languages & frameworks

Production timeline

2025

RaySuite AI

Senior Software Engineer

CURRENT

Jun 2025 – Present · Delhi, India (Hybrid)

focus → AI + infrastructure

  • Agentic AI
  • Campaign Automation
  • Predictive Analytics
  • Microservices
  • Distributed Databases
deployment log
  • Developed an agentic AI system, Adi, using LangGraph architecture for intelligent workflow automation.
  • Built scalable microservices for managing Meta and Google ad campaigns.
  • Created AI-powered chatbot systems for automated social media engagement and quick replies.
  • Designed predictive algorithms to forecast ROAS and revenue for brands.
  • Implemented automation that optimises campaign budgets dynamically based on performance.
  • Automated generation of static and video ad copies within campaigns.
  • Managed backend infrastructure and databases, including MongoDB, Cassandra and PostgreSQL.
2023

Nitro Commerce

Software Engineer (DevOps)

SHIPPED

Nov 2023 – May 2025 · Gurugram, India (Hybrid)

focus → DevOps + cloud infrastructure

  • Kubernetes
  • 100K+ RPM infrastructure
  • CI/CD
  • Grafana + Loki
  • GCP / AWS
deployment log
  • Independently managed production Kubernetes infrastructure serving applications handling 100K+ requests per minute, with load balancing.
  • Implemented centralised monitoring and log management using Grafana and Loki.
  • Designed and maintained GitHub Actions CI/CD pipelines, reducing deployment time and enabling zero-touch production releases.
  • Managed 20+ production Linux servers across GCP and AWS.
  • Configured and maintained MongoDB, Cassandra and PostgreSQL.
  • Set up and managed onsite infrastructure and networking for Loyalty Rewardz.
2022

Wigzo by Shiprocket

Associate Software Engineer

SHIPPED

Aug 2022 – Oct 2023 · Delhi, India

focus → backend engineering

  • Backend Engineering
  • Production Releases
  • Database Management
  • Client Issue Resolution
deployment log
  • Contributed across the software development lifecycle: coding, testing and bug fixing.
  • Prepared and deployed software release components.
  • Maintained and optimised databases and application workflows.
  • Worked directly with clients to resolve SMS and WhatsApp campaign and job execution issues.
  • Worked on MVC-based systems and backend services.
2022

K.R. Mangalam University

B.Tech, Computer Science Engineering

Graduated 2022

Infrastructure × AI

At RaySuite AI, the infrastructure work now runs underneath an agentic system.

  • Agentic AI system“Adi”, built on LangGraph for intelligent workflow automation.
  • Campaign automationMicroservices managing Meta and Google ad campaigns, with budgets optimised dynamically from performance.
  • Predictive analyticsAlgorithms forecasting ROAS and revenue for brands.
  • AI-powered chatbotsAutomated social media engagement and quick replies.
  • Ad generationStatic and video ad copy generated inside campaigns.
  • Data and servicesMongoDB, Cassandra and PostgreSQL behind the platform.

Selected projects

01 / KUBERNETES

Production Kubernetes Infrastructure

Independently ran production clusters behind a load balancer, serving 100K+ requests per minute across GCP and AWS servers.

  1. Internet
  2. Load Balancer
  3. Kubernetes
  4. Microservices
  5. Databases
  • Kubernetes
  • GCP
  • Docker
  • NGINX
  • Linux

02 / CI/CD

CI/CD Automation

GitHub Actions pipelines that cut deployment time and made production releases zero-touch.

  1. GitHub
  2. GitHub Actions
  3. Docker Build
  4. Testing
  5. Production
  • GitHub Actions
  • Docker
  • CI/CD
  • Automation

03 / OBSERVABILITY

Observability Platform

Centralised monitoring and log management with Grafana and Loki, plus log retention on Google Cloud Storage.

  1. Applications
  2. Logs
  3. Loki
  4. Grafana
  5. GCS
  • Grafana
  • Loki
  • GCS
  • Monitoring
  • Logging

04 / SECURITY

SSL / Reverse Proxy Automation

Reverse proxy and certificate automation using Certbot and Kubernetes cert-manager with Let's Encrypt.

  1. Client
  2. NGINX
  3. SSL / Cert-manager
  4. Kubernetes Service
  • NGINX
  • Cert-manager
  • Let's Encrypt
  • Kubernetes

Engineering principles

  1. 01

    Automate the repeatable

    If it has been done by hand twice, it becomes a pipeline.

  2. 02

    Observe before you debug

    Logs, metrics and dashboards first. Guessing is slower.

  3. 03

    Design for failure

    Nodes die and pods restart. The system should barely notice.

  4. 04

    Keep deployments boring

    Zero-touch releases mean nobody holds their breath on release day.

  5. 05

    Scale the system, not the problem

    Find the bottleneck first. Then add capacity.

Establish connection

Open to conversations about infrastructure, platform engineering and AI-driven automation.

CONNECTION REQUEST STATUS: WAITING