Comprehensive car auction platform built with Flutter. Features live bidding, multi-language support, video streaming with Tencent VOD, and about 37,000+ car model database. Currently in beta testing.
Comprehensive car auction platform built with Flutter. Features live bidding, multi-language support, video streaming with Tencent VOD, and about 37,000+ car model database. Currently in beta testing.
Bidauto is a comprehensive car auction and trading platform built with Flutter, enabling users to buy, sell, and bid on vehicles through a modern, intuitive interface with UI/UX quality comparable to industry leaders like eBay. Supporting iOS, Android, and Web platforms, Bidauto provides live bidding, video content management, and detailed search capabilities for the automotive marketplace.
This hybrid FlutterFlow and custom Flutter project features efficient handling of a massive ~37,000+ car model database with detailed filtering options allowing users to search by specific car make, model, generation, location, lot type, and more. The platform supports multiple languages (English and Traditional Chinese) and integrates with Firebase for push notifications that work seamlessly across mobile and web platforms.
The application includes media management with Tencent Cloud VOD integration for video streaming with standard compression, custom video player controls, multi-image capture capabilities, and interactive car damage visualization. With features like deep linking and location-based search, Bidauto provides a complete solution for the modern automotive marketplace.
**Note:** Development completed after 6 months. The application is technically ready and in beta testing, but public launch is pending business-side preparation (marketing, user acquisition). Not publicly available on App Store or Google Play Store yet.
Managing ~37,000+ car model database: bundle vs download tradeoff
Client suggested downloading the database from server to reduce app size, but I opted to bundle it as a ~23 MB JSON asset to guarantee offline access and avoid network dependency on first launch. Implemented efficient parsing with lazy-loading and caching of frequently accessed models. While this increased initial download size, it ensured instant availability for the auction use case where speed is critical. In retrospect, a hybrid approach (SQLite + progressive download) would have been optimal.
Slow video playback and unresponsive seeking compared to platforms like YouTube
Identified that raw video files were causing poor playback performance. Took initiative to research solutions and discovered Tencent Cloud VOD service with adaptive bitrate streaming. Integrated Tencent VOD SDK for Flutter, implementing video upload, transcoding, and optimized playback with automatic thumbnail generation. Added YouTube-like seek preview feature using 5-segment sprite sheets, allowing users to preview content at 1/5, 2/5, 3/5, 4/5, and 5/5 positions while scrubbing. This resulted in 8x faster playback speed and near-instant seeking. Subsequently guided the web engineer through implementing the web SDK to achieve consistent video performance across all platforms.
Version 1.0.87 successfully deployed to beta testing
Supporting iOS, Android, and Web platforms
8x improvement in video playback performance
Efficient handling of 37,000+ car model database
Let's discuss how I can help with your similar requirements.
Get in Touch