AutoSphere - NextGen Automotive Marketplace in Rwanda
Project Overview
Category
E-Commerce Platform
Client
James Wilson
Timeframe
Q4 2022 - Q2 2023
Role
Lead Developer
Goals
- Enable real-time vehicle configuration with dynamic pricing
- Reduce search-to-listing time by 40%
- Increase dealer adoption by 35% within the first 90 days
How I Grew
This project accelerated technical expertise: - Gained proficiency in 3D visualization and WebGL through Three.js. - Developed client-side machine learning models for dynamic pricing. - Mastered optimization of complex data structures for real-time updates. - Enhanced AWS deployment strategies for global scalability.
Challenges
The automotive sector posed complex challenges: - Managing extensive vehicle datasets with hundreds of attributes per listing. - Delivering immersive 3D visualizations without compromising performance. - Unifying B2C and B2B functionalities in a single platform, unlike existing solutions that prioritized one over the other.
Solution
- Engineered a unified data model to streamline consumer and dealer interactions. - Implemented interactive 360° vehicle views with hotspot annotations using Three.js. - Developed a TensorFlow.js-based machine learning model for fair pricing recommendations. - Built a dealer portal with inventory management and lead tracking capabilities. - Designed a progressive loading system to optimize media-heavy pages, ensuring responsiveness.
Results
- Reduced search-to-listing time by 52%, exceeding the 40% goal. - Achieved 68% dealer adoption within 90 days, surpassing the 35% target. - Attained a 4.8/5 average rating for search relevance. - Processed $28M in transactions in the first quarter post-launch.