AWS & Azure · Principal-level architecture

Your cloud bill is an architecture problem. I find it in seven days.

I’m a principal cloud architect with 11 years across AWS and Azure. Fixed-scope audits that hand your engineering team a prioritised, code-level remediation plan — not a slide deck of recommendations.

Book a 20-min discovery call See the five engagements

11^yrs

Architecture & platform engineering

Dual

Hands-on across AWS and Azure, not one plus theory

7^days

Typical turnaround on a FinOps audit

250+

Published technical guides on AWS & DevOps

01 — The premise

Cloud waste is almost never a billing problem. It is an architecture decision that nobody has revisited since the last funding round.

Cost Explorer and Azure Advisor will tell you an instance is idle. They will not tell you that your NAT Gateway egress is cross-AZ by accident, that your queue workers should be Spot, or that a Graviton migration would pay for the whole quarter. That judgement needs someone who has actually run the migration — which is the entire basis of how I work below.

02 — Engagements

Five fixed-scope engagements. No open-ended retainers you cannot forecast.

Most requested

FinOps & Cloud Cost Audit

For teams spending $3,000+/month on AWS or Azure.

from $1,200

A complete inventory sweep across every account and subscription, then a ranked remediation roadmap separating the quick wins from the structural work.

Orphaned volumes, unattached IPs, oversized instances, idle databases
Graviton / ARM migration and right-sizing evaluation
Reserved Instances vs Savings Plans vs Spot commitment modelling
S3 / Blob lifecycle tiering and snapshot cleanup policy
Executive summary plus a line-by-line engineering backlog

Request a FinOps audit →

Well-Architected & DevSecOps Review

Teams scaling, raising, or chasing SOC2 / ISO.

from $1,800

A structured pass over security posture, resilience and pipeline integrity, ending in a risk matrix with remediation snippets you can merge.

Least-privilege IAM review, MFA/SSO enforcement, secret management
Multi-AZ configuration, failover testing, RDS/NoSQL bottleneck analysis
SAST/DAST and container scanning inside the pipeline
Risk matrix graded Critical / High / Medium with Terraform fixes

Production Observability Setup

Teams finding out about outages from customers.

from $2,000

Full-stack visibility built on the four golden signals, with alerting tuned so the pager means something.

Prometheus/Grafana, OpenTelemetry, CloudWatch or Azure Monitor
Latency, traffic, errors and saturation dashboards
SLO and burn-rate alerting routed to Slack or PagerDuty
Centralised log aggregation and structured tracing

Fractional Cloud Architect

Founders and CTOs who need the judgement, not the headcount.

from $1,500 / mo

Standing architectural oversight on retainer — roadmap input, design review and escalation cover.

Fortnightly architecture review calls, 10–15 hours a month
Asynchronous RFC and pull request design review
Pre-migration and vendor sizing consultation
Priority access for critical incident post-mortems

Technical Interviewer-as-a-Service

Companies and agencies hiring cloud and DevOps engineers.

from $150 / candidate

An active principal architect runs the technical round, so your engineering leadership does not lose 40 hours a hire.

60-minute deep dive: scenario troubleshooting, system design, IaC
Objective scorecard delivered within 12 hours
Strong Hire / Hire / No Hire with written technical rationale
Recorded session notes plus a strengths and gaps analysis

03 — Selected work

Problem, intervention, measured outcome.

Client names withheld under NDA. Architecture and tooling described exactly as built.

Azure AKS → AWS EKS

Cross-cloud microservices cutover with no planned downtime

Zero downtime

Across the production cutover

Problem

A nine-service .NET Core backend on Azure AKS hit persistent database connection throttling, with slow release deployments and awkward multi-region failover.

Intervention

Modular Terraform and Helm provisioning of the EKS target, dual-ingress split routing via Route 53 weighted records for a gradual canary shift, and state validation before legacy node pools were drained.

Outcome

No data loss and no unplanned downtime across the production database cutover, with Karpenter handling autoscaling on the EKS side afterwards.

Terraform AWS EKS Azure AKS Helm Karpenter Route 53

Azure + AWS hybrid

Production GenAI document intelligence platform, built greenfield

Reproducible IaC

No manual console drift

Problem

A client needed structured insight out of unstructured PDFs and scans, with an event-driven workflow that kept both latency and compute cost under control.

Intervention

Entire environment provisioned from scratch in Terraform via Azure DevOps. Azure AI Foundry with Azure PostgreSQL pgvector for RAG retrieval, a hybrid OCR pipeline on App Services paired with Adobe Extract, and SQS + EventBridge for asynchronous multi-stage ingestion.

Outcome

A reproducible, end-to-end RAG architecture deployed with no manual console drift.

Azure AI Foundry pgvector Terraform SQS EventBridge Azure DevOps

Elastic Beanstalk → ECS

Breaking a PaaS monolith into properly packed containers

Denser packing

Zero-downtime rolling updates

Problem

Running microservices on Elastic Beanstalk forced an oversized EC2 instance per application, giving poor bin-packing and little deployment flexibility.

Intervention

Dockerised the services, decoupled shared dependencies, and moved workloads to ECS on Fargate and EC2 launch types behind path-based ALB routing with task-level IAM roles.

Outcome

Markedly better compute density per dollar, and zero-downtime rolling updates in place of whole-environment swaps.

AWS ECS Fargate Docker ALB IAM

04 — For hiring teams

Hiring a senior cloud engineer? Have one vet the candidate.

A 60-minute technical deep dive covering live scenario troubleshooting, system design and IaC reasoning — the things a CV cannot tell you. You receive a scored, written recommendation within 12 hours.

Strong Hire / Hire / No Hire

With written technical rationale, not a vibe check.

12-hour scorecard

Strengths, gaps and the exact questions asked.

Per-candidate pricing

from $150 per candidate

Book candidate vetting

05 — How an audit runs

Four steps, and you can stop after any of them.

Discovery call

Twenty minutes. What you run, what it costs, what hurts. No obligation beyond the call.

Scope and NDA

A written statement of work with a fixed fee, and a mutual NDA signed before any access.

Read-only audit

A cross-account read-only IAM role. Every query is visible to you in CloudTrail.

Report and walkthrough

Executive summary, ranked engineering backlog, and a live session with your team.

06 — Indicative pricing

Published, so you can budget before we speak.

USD INR

Engagement	From	Billing
FinOps & Cloud Cost Audit	$1,200 – $2,500	Fixed fee, one sprint
Well-Architected & DevSecOps Review	$1,800 – $3,500	Fixed fee, one sprint
Observability Implementation	$2,000 – $4,000	Per project
Fractional Cloud Architect	$1,500 – $3,000 / mo	10–15 hrs/mo, 2-month minimum
Technical Interviewer-as-a-Service	$150 – $250	Per interview session

Scope is agreed in writing before any access is granted. If a discovery call shows an audit would not pay for itself, I will say so on the call.

07 — Writing

Two hundred and fifty technical guides, written over four years.

EKS cluster builds, Lambda and SQS cost behaviour, CloudFormation patterns, Kubernetes troubleshooting. It is the most honest evidence of how I think about infrastructure — read a few before you book anything.

Browse all writing →