Specialty Sales Classics

The Specialty Sales team knows a thing or two about vintage and classic cars - they’ve been buying and selling the world’s best classic cars since 1978.

Behind the showroom runs their incredible DMS (Dealer Management System), a proprietary platform that allows their team to efficiently manage all aspects of their business, including vehicle consignment, marketing, sales, customer payments, and much, much more. It is the system the entire operation lives in, managing the whole vehicle lifecycle from the day a car is consigned to the day it drives off to a new owner.

Specialty Sales came to us looking for help after parting ways with a development firm that had done a major disservice to their codebase. Years of poorly written software had accumulated, and the standards that keep a codebase healthy had been neglected for just as long. The result was a familiar and painful pattern: the team’s time and budget went into resolving bugs and working around problems instead of into making the business better.

A rescue like this comes with a hard constraint. The DMS could not be paused while its software was repaired. Consignments, sales, and customer payments had to keep flowing every single day, which meant every fix had to land in a live system that the business depended on. Beneath the immediate bug-fixing sat a larger goal: bring the platform to a state where it stopped consuming the company’s attention and started earning it.

We see this situation often enough to know how it develops. A business commissions custom software in good faith, the firm building it cuts corners the client has no way to see, and by the time the symptoms surface, the relationship has soured and the codebase carries years of compounding shortcuts. The software still ran the business - it just made every week harder than it needed to be.

We quickly set to work fixing years of poorly written software and implementing long-neglected standards. The order of operations matters in a rescue: stabilize the worst problems first, establish test coverage so each fix stays fixed, and only then shift attention to improvements. Every bug resolved under that discipline is resolved once, instead of returning in a new form two months later.

The DMS is a Ruby on Rails application with an Ember.js frontend, integrated with DocSpring for document generation and the payment processor the business already used, running on PostgreSQL and AWS infrastructure. Those integrations set the bar for the quality work: a system that takes customer payments and produces the documents a dealership runs on has no room for the casual defects the codebase had been carrying. As the codebase recovered we did the deeper work too: redesigning what needed it, carrying the platform through Rails upgrades so it never drifted onto unsupported versions, migrating infrastructure to the cloud, and putting continuous integration and delivery in place so the growing RSpec suite runs before anything ships.

Over time, the effort spent on resolving bugs and other problems gave way to time spent adding features and improving their software. That crossover is the milestone that matters most in a rescue engagement.

With increased efficiency, flexibility, and resource availability gained from refinements to their internal operations, we had the opportunity to rebuild and modernize their customer-facing application.

The public website is where classic car buyers meet the inventory, and the old interface had grown outdated and cluttered. With a simple goal of transforming it, we redesigned the site with a modern touch and overhauled what sat behind it, bringing their front-facing code up to industry standards and the same automation the internal platform now enjoys. The consumer product and the DMS stopped being two codebases in two different decades and became one platform held to one standard.

Fast forward to today: their platform has been redesigned, upgraded, has a solid base of test coverage, and is essentially bug free. The team that once budgeted its weeks around software problems now runs its consignment, sales, marketing, and payment operations on a system that simply works, and development effort goes toward whatever the business wants next rather than toward the backlog of the past. The measure of a rescue is whether the client ever has to think about the old problems again, and here the answer has been no for years.

This continuous effort and constant support has kept their entire platform, both the internal DMS and the public-facing website, up-to-date and ready for the company’s next business move. It is the outcome we aim for in every rescue: a repaired codebase that has grown into a platform the business can plan on.