Full disclosure: I'm the CIO at Great North Airlines. So when I say "we built this website," I mean I stayed up way too late convincing Claude to write better TypeScript while juggling my actual airline responsibilities.
This is the story of how we went from "our Squarespace looks like a vacation booking site" to "holy crap, we actually look like a real aviation company" in about three weeks.
The Problem Nobody Wanted to Admit
Our old site was embarrassing. Not "a little dated" embarrassing. Like, "we're trying to win a M government contract and our website has the same template as a yoga studio" embarrassing.
Here's what actually happened: We'd been pitching ACMI contracts to provincial governments and corporate travel managers. The conversations would go great until someone said "let me check out your website." Then silence. Then a polite email about "going in a different direction."
The reality is, if you're asking a government procurement officer to trust you with emergency medevac operations, your website can't look like you made it during a free Squarespace trial.
Why I Didn't Just Hire an Agency
I looked at agencies. The quotes came back at 0K-80K with 4-6 month timelines.
For context, I'm running IT for an airline that operates CRJ-200s and Dash 8s across Northern Canada. We don't have agency budgets. We have "figure it out" budgets.
Plus, I knew exactly what we needed. I didn't need three rounds of discovery meetings to tell someone that aviation B2B is different from consumer travel. I live this stuff.
The Actual Build (What Worked and What Was Messy)
I used Claude Code for about 80% of the development. Here's the honest breakdown:
What worked stupidly well:
- Next.js page scaffolding. I'd describe a service page structure and get working code in minutes.
- Component consistency. Once I established the design system, Claude kept things uniform across 15+ pages.
- Azure deployment configs. Static web apps, CDN setup, custom domains—all generated correctly on first try.
What was messy:
- The AI-generated aircraft illustrations took way more iteration than expected. Turns out "professional aviation illustration" and "AI image generator" don't speak the same language. I probably regenerated those CRJ images 40 times.
- Mobile responsiveness needed manual tweaking. AI got it 70% right, but that last 30% was me at 11pm pixel-pushing.
- SEO content kept drifting into corporate speak. I had to actively fight against "leverage our comprehensive solutions" sneaking into every paragraph.
Timeline reality:
- Week 1: Core architecture, design system, homepage
- Week 2: All six service pages, fleet section, contact forms
- Week 3: Content refinement, mobile fixes, staging feedback, launch
Three weeks. Not three months. For a fully custom Next.js site on Azure with professional imagery.
What We Actually Built
Six dedicated service pages because "we do charter stuff" doesn't win contracts:
- ACMI (wet lease) - for operators who need aircraft + crew
- Emergency Operations - medevac, wildfire support, disaster response
- Contract Charter - scheduled service for mining camps, remote communities
- FIFO - fly-in/fly-out for resource sector workers
- Ad-hoc Charter - one-off flights for corporate or government
- Aircraft Management - for owners who want us operating their planes
Each page explains what it is, who actually uses it, and how to start a conversation. No fluff.
Fleet presentation that doesn't suck. Specs, configurations, range maps. The stuff a procurement officer actually needs to check boxes on their RFP evaluation.
Mobile-first everything. 43% of our traffic is mobile, and government networks are notoriously slow. We optimized aggressively—pages load in under 2 seconds on 3G.
The Uncomfortable Truth About AI Development
I'm going to say something that might sound like I'm undermining my own consulting business: AI didn't replace expertise. It replaced typing.
I still needed to know that government aviation procurement cares about TC certification status. I still needed to understand that ACMI clients want to see crew utilization rates, not marketing photos of smiling passengers. I still needed 15 years of enterprise IT experience to architect a solution that actually works.
What AI did was let me execute at 10x speed once I knew what to build.
The uncomfortable corollary: if you don't know what you're building, AI just helps you build the wrong thing faster.
Results (So Far)
It's been two months since launch. Here's what's changed:
- RFP response rate is up. Procurement officers actually read our proposals now instead of bouncing after checking the website.
- The rebrand to Great North happened smoothly. New domain, new identity, zero technical hiccups.
- I've gotten three calls asking "who built your website" from other aviation operators. (Hi, that's what ForIT does.)
Is it perfect? No. I still see things I want to fix every time I visit. The About page could use more personality. The contact form success message is boring. The illustrations still look slightly AI-generated if you squint.
But it's live. It's working. And it cost a fraction of what an agency would have charged.
What This Means for Aviation Companies
Here's my pitch, and it's a selfish one because I consult in this space:
Aviation is stuck between two bad options. Enterprise software that costs 00K and takes 18 months to implement. Or consumer tools that don't understand the industry.
AI-assisted development is the third option. Move fast, stay cheap, but bring actual domain knowledge so you're not just building generic nonsense with aviation keywords sprinkled in.
I built my own airline's website this way. I know it works because I use the result every day.
Visit Great North Airlines at greatnorthairlines.com. If you want to talk about what AI development could do for your aviation operation, let's chat.