Hank Karpinen

Full-stack engineer · Pullman, WA · open to roles

I rebuild the
software nobody
wants to touch.

Six years ago I was a manual tester writing bug reports. Today 16,000 students, faculty and staff use systems I rebuilt from scratch.

The 30-second version

What that actually amounted to.

students, faculty and staff use software I built or rebuilt

cheaper to run, every year, after one platform rebuild

separate legacy systems replaced by one application

less code left behind in every app I modernized

AWS Solutions ArchitectAWS Cloud PractitionerCertified ScrumMaster.NET · React · AWS · Postgres

Before → after

Three things
measurably changed.

Cost to run the college's website

A Drupal 11 rebuild that shipped more features, faster.

old platform
−80% / yr

Code to maintain per app

Same features, a quarter of the places a bug can hide.

legacy app
−75%

Time to design a system rewrite

I built an AI tool that does the slowest part of the job.

months to yearsdays

How I got here

Four jobs. One consistent
job description.

  1. 012020—21

    Learned how software fails

    Manual tester at QualityLogic. A year spent watching exactly how shipped software breaks.

    QualityLogic · Boise, ID

  2. 022022—24

    Automated it away

    100+ automated tests against live environments for a national consumer brand — and a front-row seat to a monolith being split apart.

    Scentsy · Meridian, ID

  3. 032024—25

    A unified experience for all of campus

    Ten disconnected systems became one app used by 1,600 people — RA logs, WiFi access, student elections. Then rebuilt the public website for 80% less.

    The College of Idaho · Caldwell, ID

  4. 042025—now

    Doing it at scale

    Modernizing decade-old applications for a university of 15,000 — the campus directory, the hazardous-chemical disposal system, a statewide study portal.

    University of Idaho · Moscow, ID

The part I'm proudest of

I didn't just do the slow work. I built the tool that removes it.

Rewriting an old system is mostly archaeology — reading code until you understand what the business really does. That phase is where these projects die. Mine reads the codebase, extracts the real business concepts, designs the architecture, and scaffolds it in whatever stack you're targeting.

  1. 01Read the legacy codebase
  2. 02Extract the real business concepts
  3. 03Design a proven architecture
  4. 04Scaffold the new system — in days

Chapter 05 · you're standing in it

So I built one more, to show you.

This site runs on a suite of apps I built end to end: shared-living tools for bills and chores, a community forum, expense tracking, and a couple of open utilities. It's all running: seven services, thirty-plus screens.

See how it works →

One click and you're inside.
No signup, no email.

The demo account resets nightly and nothing you do persists. You'll be in within seconds.

Find the system everyone works around instead of with. Replace it without the business going dark.

Ships end to end, alone

Interface, API, database, cloud, deployment. This whole website is the proof.

Deploys in minutes, hands-off

One merge builds, ships and scales itself. Releases stopped being an event.

Starts from how it breaks

Four years in QA first. Reliability is where I begin, not what I patch later.

Talks to the people affected

Certified ScrumMaster. I scope with stakeholders and run the ceremonies.

Got a system nobody
wants to touch?

Mid-level and senior full-stack roles · Remote, hybrid or onsite · Pullman, WA