GMC Extractor — AI News & Media Platform
Qualitative news article AI intelligence and media monitoring desktop workstation shipped for automated digital press archiving, keyword alerting, and analysis.
Why this was built
Monitoring multi-channel national news publications manually was overwhelming research teams. Analysts spent hours reading, sorting, and transcribing articles across disparate media websites, leading to transcription delays and missed high-priority alerts.
Technical Approach
Engineered a native cross-platform desktop application using Flutter Desktop for Linux, macOS, and Windows. Built an automated HTML scraper and media extraction engine that parses articles, applies AI-driven qualitative categorization, cross-references Excel guidelines, and syncs cleanly to Supabase cloud and local SQLite storage.
// GMC Extractor: Global HTTP overrides and extraction service setup
class AppHttpOverrides extends HttpOverrides {
@override
HttpClient createHttpClient(SecurityContext? context) {
return super.createHttpClient(context)
..badCertificateCallback = (X509Certificate cert, String host, int port) => true;
}
}
// Extraction provider invoking AI intelligence and SQLite archive
Future<void> runMediaExtraction(OutletSource outlet) async {
final rawHtml = await scraperService.fetchOutletFeed(outlet.url);
final extractedArticle = await aiParser.extractQualitativeData(rawHtml);
await databaseService.saveArticle(extractedArticle);
await supabaseSyncService.enqueueForCloudReconciliation(extractedArticle);
}Key Architectural Features & Outcomes
Cross-platform compilation across Linux, macOS, and Windows
Automated web scraping and HTML parsing for major digital outlets
AI-assisted qualitative categorization and article intelligence
Local SQLite persistence with offline-first Supabase cloud synchronization
Multi-format media format support (video, audio, text archives)