Hi, I'm Cory.

I build the software dealers use to price the cars they take in on trade.

That means React front ends, .NET services, Terraform, the database underneath, and whatever is wrong with the build this week. Five years in, I've gone from writing the tests to owning the systems.

I didn't start in software. I spent ten years in quality at a plastics manufacturer, from the R&D lab to running a team of eight engineers. Along the way I started building the tools we used, first in SharePoint and Excel, then an incident-reporting system and document control software that replaced manual approvals. In 2020 I taught myself to code properly and started over. I joined as a test engineer in 2021, built the team's automation framework from an empty repository, and moved into engineering the following year.

I also run a small development studio, Violet Summit Forge (opens in a new tab), which keeps me honest about what it takes to ship something end to end for someone who is paying.

How I work

I read other people's code for a living

About 700 reviews so far, for engineers from their first month to senior. I once declined to approve a redesign that spanned three repositories. Going through it I found three problems with the design, and a bug that was already live in production, which the new design would have rebuilt. Telling people “not yet” on work they've put weeks into is never fun, so I bring what I found with me.

I like being able to turn things off

When I added a new capability to the appraisal workflow, I chained its feature flag behind an existing one, so the whole thing can be switched off from a single place. Shipping is the easy half. I want to know how I'd take it back before I ship it.

I want to be able to follow a request

On the wholesale platform I added correlation IDs across four services so a single request could be followed from end to end, and I moved those services onto shared CI workflows so one fix reached all of them. Later I took on SRE duty for a reporting component that dealers use across the country. Nothing teaches you what a system should tell you when it's unhappy like being the one who gets the page.

I go looking for the root cause

Ten years of 8D, Six Sigma and Kaizen left me with a habit: before I say something works, or that I fixed it, I find out why. When I moved a production database tier onto cross-region disaster recovery, I wrote the enablement spec and the validation plan, and fixed the two Terraform failures that were blocking the secondary cluster. The plan is how I know it works.

I sit with people until it clicks

I've mentored nine engineers, from entry level to senior, and I've advised people up to director level. The part I like is when someone stops needing me for the thing we were stuck on.

Recent work

Shipped a change that crossed three repositories and multiple release trains (an API, a BFF layer and a React micro-frontend) to make vehicle cost evaluation and option selection more robust.
Cleaned up CI security across nine repositories in two GitHub organizations: OIDC for artifact authentication, and up-to-date static analysis everywhere.
Migrated a standalone reporting product into the core application and retired the old system.
Helped build a B2B wholesale platform from scratch. It publicly reported over 350,000 bids and 15,500 vehicles sold in its first fourteen months.
Built internal AI tooling that my team uses day to day: reusable agent skills for discovery, review and release, and a searchable code knowledge graph.

What I work with

Languages: C#/.NET, TypeScript, JavaScript, SQL, Terraform (HCL)
Front end: React, Redux, Stencil.js, micro-frontends, Next.js
Back end: .NET, REST APIs, BFF architecture
Data: PostgreSQL, Oracle, Aurora, MySQL, MongoDB
Cloud: AWS, Terraform, Kubernetes, ECS, cross-region disaster recovery
Reliability: New Relic, Datadog, PagerDuty, SLI/SLO design, on-call
Delivery: GitHub Actions, SonarQube, Artifactory, OIDC, feature flags

Bring me in when

A platform has to move without breaking the people using it: a migration, a consolidation, a disaster-recovery cutover.
CI and supply-chain security need to go from “mostly fine” to something you can defend.
Someone needs to ask the uncomfortable question about a design before the code gets written.
A front end, a service and the infrastructure under them all need to line up.
A team of mixed experience needs someone who makes everyone around them better.

If one of those sounds like your week, let's talk.