---
title: "Senior Developer"
author: "Jarrod Carlson"
company: "Trineo Ltd"
industry: "Software consulting"
location: "Boulder, CO"
period: "Oct 2016 – Sep 2018"
websites:
  - "https://www.ikonpass.com"
  - "https://clearchanneloutdoor.com"
tech:
  - "Ruby on Rails"
  - "React"
  - "PostgreSQL"
  - "Heroku"
  - "Node.js"
  - "Capybara"
canonical: "https://builtbyjarrod.com/experience/trineo"
---

# Senior Developer — Trineo Ltd

Oct 2016 – Sep 2018 · Boulder, CO

> New Zealand-born software consultancy and Heroku partner, later acquired by Salesforce

**Summary:** Shipped client products for a consultancy later acquired by Salesforce — the Ikon Pass e-commerce storefront for Alterra, and the proof-of-performance platform that retired Clear Channel Outdoor's century-old Polaroid-and-mail workflow.

---

Trineo was a software consultancy headquartered in Christchurch, New Zealand — a
longtime Heroku specialist that Salesforce would eventually acquire outright.
They hired me out of Atlanta and relocated me to their brand-new Boulder office,
established to anchor their growing US client base. Working as the senior
US-based engineer alongside teammates in New Zealand and Norway, our follow-the-sun
setup covered clients nearly 20 hours a day — without anyone pulling
overnight on-call duty.

## The Ikon Pass storefront

Trineo had worked with ski resort operator Intrawest for years — through its
acquisition by KSL Capital Partners and rebranding as Alterra Mountain Company —
so work on the ski pass platform was underway before Alterra or the Ikon Pass
had even been publicly announced. The solution was built from scratch: a Ruby on Rails
backend, React frontend, and PostgreSQL, all running on Heroku, where seasonal
on-sale traffic spikes were something the platform absorbed rather than a crisis
we engineered around.

I joined the project team for the Ikon Pass storefront's first sales season — the phase
where integration reality hits. A ski pass isn't a simple cart checkout; it's an
identity product with a physical footprint. Customers who bought by phone had to
complete their profiles online, upload the headshot printed on their pass, and
sign electronic waivers. Online purchases had to be visible instantly to
in-person customer service — and vice versa. And every pass had to sync out to
each resort's entry hardware, which ranged from modern RFID gates to simple
offline barcode scanners, each with its own integration quirks. I was the sole
US-based engineer on the team (alongside three or four in New Zealand and one in
Norway, plus a product manager and designer), making the technical calls and
building the automated integration test suite that bridged the Rails backend and
the Node rendering layer with Capybara.

## Proof of performance, minus the Polaroids

Billboard advertising runs on Proof of Performance: evidence for the client that
their ad actually went up, where and when it was supposed to. For roughly a
century, that meant someone driving around the city with a camera, photographing
billboards, filling out paper forms, and mailing the packet to the advertiser.

For Clear Channel Outdoor, we replaced that workflow end to end. We built the
mobile-friendly capture app field crews used to photograph billboards —
geo-tagged and timestamped, so the photo itself was the proof — and the client
portal where advertisers logged in to see every ad buy on a map, filterable by
campaign and run dates. The hundreds of labor hours saved every month were all
the hours no longer spent printing, packaging, and mailing proof to clients.

## Consulting is a contact sport

Our product owner framed *what* to build; much of *how* — and *by when* — fell
to me. A regular part of my job was sitting in client meetings translating
feature requests into commitments everyone could live with: when a client wanted
X in Y days and that math didn't work, negotiating the scope down or the
timeline out until it did.

Two years of shipping client products on Heroku also hardened a conviction I'd
been forming for a while: hosting environments are commoditized and
non-value-add, and teams that specialize in business software shouldn't be
hand-rolling infrastructure. That conviction became the blueprint for the
[platform-as-a-service I'd go on to build at Cox Automotive](/experience/cox-automotive.md).