Newspresso Tech

DevOps, full-stack, and AI-native engineering, from a team that operates its own product.

We run Kubernetes clusters, cut cloud bills, ship web and mobile products, and build the AI systems behind them, end to end, the same practices that keep Newspresso, our own news app, running in production every day.

Book a call See the work

About

A consultancy run by the people who build the product, not just advise on it.

Newspresso Tech is the engineering practice behind Newspresso, our own news app, brought to outside clients as a small engineering team based in India, working directly inside your time zone.

DevOps & Cloud Engineer | Founder at Newspresso Tech

Watch on YouTube ↗

The one value we hire, work, and grow around

Ownership.

We don’t run a long list of “core values.” We focus on one, because we believe ownership is upstream of the rest of them: someone who genuinely owns the outcome is already being responsible, already bringing real value to the work, and already growing into more than the job description asked for.

Responsibility, not tickets

Ownership means noticing what's broken and fixing it, not waiting to be assigned a ticket for it.

Growth, not just output

The people who take ownership are the ones we invest in developing further, technically and professionally, within the company.

Trust, not a sales layer

For clients, ownership means direct accountability: the person who owns your problem is the person doing the work.

01 / WHAT WE DO

Three things. Done properly.

We specialize narrowly on purpose: DevOps, full-stack, and AI-native delivery, not a little bit of everything.

DevOps & Platform Engineering

Kubernetes, cloud cost, and reliability, run by people who carry the pager.

Lower cloud spend without a reliability trade-off
Clusters that pass audits instead of triggering them
CI/CD pipelines your team can actually maintain

Learn more

Full-Stack Development

Web and mobile apps, built and operated end to end, like we do for our own product.

One team across frontend, backend, and infrastructure, no handoff gaps
An SEO-first web presence built to be found, not just to look good
A mobile app shipped to both the App Store and Google Play

Learn more

AI Native Engineering

AI features and pipelines, built by people who ship them into a live product.

Ship an AI feature that's demoable in weeks, not quarters
An LLM pipeline with real cost and quality guardrails, not a raw API call in prod
Search and Q&A that actually understands your content, not just keyword match

Learn more

Companies

Here are some of the companies our team has experience working at

Success stories

Explore success stories

Newspresso

Our AI-native product — designed, built, and marketed end to end by us

4.9★

Average user rating

AI-native

Content published by AI Agents

Vercel

Auto-deployed on every push

View Newspresso’s success story

LexisNexis

Platform engineering inside a FedRAMP-aligned, multi-tenant Kubernetes environment: cost, resiliency, and AI infrastructure, not a green-field project.

13%

EKS cost reduction

~8%

Lower monthly infra cost

12 hrs

Full cluster recovery

View LexisNexis’s success story

Arista Networks

A CI/CD re-architecture: consolidating a sprawling Jenkins pipeline estate into a modular, Spinnaker-driven deployment flow.

15 → 3

Jenkins pipelines consolidated

Jenkins pipelines automated

~38%

Faster dev/test toolchain

View Arista Networks’s success story

Camping World

AI-accelerated delivery: Claude-authored Harness pipelines, feature-flagged releases, and DevRev AI Agents triaging support tickets automatically.

23%

Efficiency improvement

Decoupled

Deployments from prod code

Automated

Support ticket triage

View Camping World’s success story

See all success stories

02 / STACK

What we actually use

The exact tools running in production for Newspresso and our clients, not a keyword list.

Kubernetes & Orchestration

Kubernetes

EKS

Helm

k9s

Kyverno

Karpenter

Istio

Velero

ArgoCD

Cloud Platforms

AWS

IAM

Lambda

EventBridge

Auto Scaling Groups

EC2

Graviton

CI/CD & Infrastructure as Code

Terraform

Jenkins

Spinnaker

Harness

GitHub Actions

Flux

GitOps

Ansible

Git

Docker

JIRA

Monitoring & Observability

Datadog

Splunk

Dynatrace

Coralogix

OpenTelemetry

Frontend

Next.js

React

TypeScript

Tailwind CSS

Mobile

Flutter

Dart

iOS

Android

Backend & Data

Java

Python

Spring Framework

REST APIs

PostgreSQL

plpgsql

Supabase

Firebase

AI Tools

Claude

OpenAI

Cursor

Copilot

Sarvam

Gemini TTS

Ollama

Vercel

DevRev

AI Agents

RAG Chatbots

Automations

People love working with us!

LinkedIn recommendation from Matt Story, Software Engineer, who managed Shubham directly at Arista Networks: 'Shubham approached his internship in the EngProd group at Arista with great enthusiasm. He showed both the desire and ability to learn new technologies and apply them to his project work.'

LinkedIn recommendation from Anshika Jain, ACCA Affiliate | EY GDS | Ex-GT, who worked with Shubham on different teams: 'It was a great experience working with the Newspresso Tech team to explore innovative marketing strategies powered by AI agents and identify opportunities to enhance the product's overall efficiency. I particularly appreciate their openness to new ideas and the collaborative approach to turning ideas to practical solutions.'

LinkedIn recommendation from Vasisht Chinta, Technical Architect - cloud, who worked with Shubham on different teams: 'Shubham is a wonderful person and an excellent professional to work with. His knowledge of AWS, Kubernetes and Containers runs deep including. DevOps spans a vast landscape of tools, technologies, and services, and he brings an exceptional combination of breadth and depth to this field.'

LinkedIn recommendation from Zhengwen (Dorothy) Xu, Software Engineer at Arista Networks, who worked with Shubham on the same team: 'I worked with Shubham during our internships at Arista, where we were both part of the EngProd team. He has a strong understanding of CI/CD and made valuable contributions to improving engineering efficiency during his internship. Beyond his technical skills, Shubham is a curious and hardworking person who is always willing to learn new things and take on new challenges. His technical knowledge and willingness to learn make him a great engineer to work with.'

LinkedIn recommendation from Lacina Zina, Kubestronaut | Principal Engineer | Architect | GPU Architect | HPC, who worked with Shubham on the same team: 'I had the pleasure of managing Shubham, and I was consistently impressed by his professionalism, dedication, and willingness to take ownership of his work. He approaches challenges thoughtfully, learns quickly, and collaborates effectively with colleagues. Shubham was a dependable member of our team whom we could trust to follow through on commitments and deliver high-quality results. Beyond his technical abilities, he brings a positive attitude and a genuine desire to grow and contribute. I truly enjoyed working with Shubham and would gladly recommend him to any organization looking for a talented, responsible, and collaborative professional. I am confident that he will be a valuable addition to any team.'

Blog

Read our blogs

Illustration of a robot mascot beside a workflow diagram connecting a webhook to HubSpot, Gmail, and Slack icons, titled 'Building an AI-Powered Lead Qualification Workflow in n8n'

Apr 9, 20269 min read

How I Built an AI-Powered Lead Qualification & Routing System in n8n (Production-Ready)

A webhook-driven n8n workflow that validates inbound leads, scores intent with a guardrailed AI agent, and routes each one — CRM deal, Slack ping, drafted email, or just a log line — based on how hot it actually is.

AI Automationn8nLead Generation

Read the post

Dark workflow diagram behind the title 'AI Powering Viral Ads: A Deep Dive into Automation — Unlock limitless creativity with intelligent workflows'

Feb 13, 20269 min read

Building an AI-Powered Ad Generation System for Businesses 💸

An end-to-end n8n workflow that turns a product name, description, and a couple of photos into a fully edited, titled, thumbnailed video ad — uploaded straight to YouTube, no manual editing.

AI Automationn8nGenerative AI

Read the post

n8n workflow diagram of a webhook flowing into a code node, an IF gate, then success and error response nodes, behind the title 'How I Built a Production-Safe Webhook Intake System in n8n'

Jan 22, 20267 min read

How I Built a Production-Safe Webhook Intake System in n8n

Why most automations fail before AI or CRMs ever get involved — and the defensive n8n webhook pattern that stops bad data at the door: validate, normalize, respond deterministically, no silent failures.

AI Automationn8nWebhooks

Read the post

Glowing 3D grid of illuminated blocks labeled 'Huge Pages', representing large contiguous memory pages

Jan 8, 20266 min read

Improve Performance of Memory-Intensive Applications on an EKS Cluster Using Huge Pages

Why memory-heavy workloads like ML model serving get unstable under load — page table overhead, TLB misses, fragmentation — and how configuring huge pages on Karpenter-provisioned EKS nodes fixes it.

KubernetesAWS EKSHuge Pages

Read the post

Diagram showing pending and unschedulable pods triggering Karpenter to provision just-in-time capacity, then consolidate it into optimized capacity

Oct 5, 20257 min read

Intelligently Scale Your Nodes in an EKS Cluster Using Karpenter: A Guide

Why Karpenter beats the vanilla Cluster Autoscaler on cost and speed, how its two core resources — EC2NodeClass and NodePool — actually work, and how to install it and watch it provision a node for a pending pod.

KubernetesKarpenterAWS EKS

Read the post

Comic panel of a ship's captain at the helm explaining that Kubernetes comes from a Greek word for helmsman or pilot, which geeks shorten to K8s

Sep 14, 20257 min read

Bootstrap Solution to Cache Heavy Container Images Using DaemonSets & Karpenter in an EKS Cluster

A build-it-yourself image-caching solution: use Karpenter to scale up a tainted node, a DaemonSet to pre-pull heavy images onto it, then untaint the node so real workloads can schedule on it instantly.

KubernetesDaemonSetKarpenter

Read the post

Isometric illustration of isolated, color-coded tenant blocks on a shared platform, ringed by shields and locks

Sep 21, 20247 min read

Setting Guardrails in a K8s Cluster in a Multi-Tenant Environment 🔐

Multi-tenant Kubernetes clusters cut infrastructure cost by sharing one cluster across teams — but without guardrails, that shared blast radius means one tenant can leak into another. Here's how to lock it down with Kyverno policies and NetworkPolicies.

KubernetesKyvernoMulti-Tenancy

Read the post

Illustration of a person pushing an oversized, overheating shipping container toward an hourglass, labeled 'Taking minutes to spin up'

Aug 31, 20246 min read

How Do I Reduce Startup Time of Heavy Pods in K8s? 🤯

Four ways to stop heavy container images from taking minutes to pull on EKS — ECR Pull Through Cache, eStargz, Spegel/kube-fledged, and Bottlerocket OS + Karpenter — with the real trade-off behind each one.

KubernetesAWS EKSImage Caching

Read the post

Isometric illustration of an Amazon EKS cluster provisioning on-demand and spot GPU node capacity

Aug 28, 20248 min read

Handling AI/ML Workloads on NVIDIA GPU Nodes in Kubernetes: A Guide

A practical walkthrough of provisioning and autoscaling NVIDIA GPU nodes on an EKS cluster with the NVIDIA GPU Operator and Karpenter, so ML engineers can train and test models without anyone hand-managing GPU infrastructure.

KubernetesAWS EKSNVIDIA GPU Operator

Read the post

See all posts

Looking for a strong team in DevOps, Full-Stack, or AI Native?

Book a short call with us to discuss the opportunity. No pitch deck, just a conversation about what you need.

Book a call