Duncan Anderson
Back to Projects

Online Travel Agency

Deal Engine

Customers were browsing thousands of flights but had no easy way to find the best deals. Great prices were buried in the noise, and conversions suffered.

01

The Problem

FlightHub's inventory had thousands of flights at any given time, but the best deals were hidden in the noise. Customers would browse, get overwhelmed, and leave. The sales team knew great prices existed but had no systematic way to surface them. Meanwhile, competitors were getting better at showing users exactly what they wanted to see.

02

What I Built

I built a system that continuously analyzes the full flight inventory and identifies which deals are genuinely goodnot just cheap, but good value relative to the route, timing, and historical pricing. The engine scores and ranks deals in real time so the best ones can be surfaced to users right when they're browsing. It turned a passive catalog into an active recommendation system.

03

The Outcome

Conversion rates went up because users started seeing deals that actually matched what they were looking for, right when it mattered. The system ran continuously without manual curation, which meant the merchandising team could focus on strategy instead of hand-picking deals every day.

Stack

Tech Used

PythonReal-Time AnalyticsRecommendation SystemsSQL

Interested in something like this for your business?

Let's Talk