Skip to content
Practice Work Approach About Get in touch

AWS consulting for critical systems · Albuquerque, NM

Cloud infrastructure for systems that can’t go down.

I’m Thomas Duquemin — an independent AWS consultant. I help teams running critical systems build cloud infrastructure that’s simple enough to understand, reliable enough to trust, and documented well enough to run without me.

How I think

Most infrastructure is more complicated than the job needs it to be.

There’s nothing wrong with an architecture that touches twenty AWS services — if the problem genuinely calls for it. Most of the time it doesn’t. A smaller, simpler design usually does the same job, costs less, and is far easier for your team to reason about when something breaks at 2 a.m.

So that’s where I start: the simplest thing that works, not the most impressive thing I could build. I won’t recommend a service you don’t need, and I won’t sell you an engagement you don’t need either. If a short conversation solves your problem, we’ll just have a short conversation.

Selected work

Some of the work.

Wildfire forecasting

36 hrs under 30 min

Forecast time, before and after the rebuild.

A wildfire-risk forecasting company ran its dead-fuel-moisture model as a single monolithic notebook — one forecast took the better part of two days. I rebuilt it as an event-driven, serverless pipeline: Step Functions for orchestration, Lambda for the processing stages, EventBridge for scheduling. The same forecast now finishes in under half an hour — fast enough to change what the company can offer its customers.

Flight logistics

Off the click-trail, onto code

A flight-logistics company’s CloudFormation had grown so piecemeal it was effectively click-ops — changes nobody could fully account for. I moved them onto Terraform and ephemeral environments, catalogued what actually existed, and left the team with a documented, version-controlled foundation they could run and extend themselves. I also built their multi-region disaster recovery, so an outage in one AWS region isn’t an outage for their customers.

Government & defense

Systems that have to be right

I’ve worked on infrastructure and data pipelines for government and defense programs, including U.S. Army predictive maintenance and national-security research. Different domain, same discipline: get it right, write it down, and don’t cut corners where it counts.

Cost reduction

The unglamorous wins

Not every job is a dramatic rebuild. For one client I mapped DNS resolver patterns across 20 AWS accounts, found the redundancy, and refactored it into a shared model — roughly $60K a year quietly recovered, and a simpler network left behind.

How I work

What working together actually looks like.

  • 01

    You work with me — start to finish

    No account managers, no junior engineers staffed to fill a bench. You get the person who does the work, for the whole engagement.

  • 02

    Calm beats heroic

    Production infrastructure is stressful enough. I diagnose before I touch anything, keep you posted in plain language, and don’t add drama to a problem that already has plenty.

  • 03

    Everything leaves as code, and as yours

    Your infrastructure ends the engagement as reviewable Terraform your team owns. I bring a library of modules built over years of this work, so you’re not paying me to reinvent a VPC — and nothing important lives only in my head.

  • 04

    I tend to do more than the brief

    If I spot something adjacent that’s going to bite you later, I’ll tell you — and often just fix it. The brief is where the work starts, not a fence around it.

About

“I’d rather build infrastructure than be the only person who understands it.”

I studied physics at Texas A&M and started out in national-security research. I found I’d rather build and run real systems than study them — so I moved into cloud engineering, helped build a data and cloud consulting firm from the ground up, and went independent as Epsilon Limited Solutions in 2024. Most people call me Duke, short for Duquemin.

Today I take on a small number of engagements at a time, focused on infrastructure for critical systems, and I architect cloud environments across more than 100 AWS accounts. I’m AWS-certified as a Solutions Architect and a Data Engineer, based in Albuquerque, and I work with clients across the Southwest and nationally.

One honest note to end on: I’m not the best cloud engineer alive, and I won’t pretend to be. What I am is steady, genuinely interested in the craft, and straight with you. I’ll tell you what I’d do, what I wouldn’t, and when you don’t actually need me.

Thomas Duquemin

Principal · Epsilon Limited Solutions LLC

Get in touch

Looking me up? Let’s actually talk.

Maybe you’ve got a specific problem. Maybe you just want a second opinion on what you’re running. Either way, email me directly — no pitch, no funnel, no discovery-call gauntlet.