Automated city events discovery platform
CityEventScout is a centralized platform for discovering local events collected from multiple online sources. It solves the problem of fragmented event information by automatically aggregating, cleaning, and organizing events from social media and web platforms into a single, user-friendly hub.
Tech stack
The system automatically collects events at configurable intervals and processes them using AI to generate concise descriptions, assign tags, and detect duplicates. All events are standardized to provide clean, high-quality information that users can easily browse and filter.
The platform allows users to discover relevant events faster without manual searching, while administrators only need to configure sources once. The system is designed to be deployment-ready and easily extensible by adding new scrapers or event sources.
The solution is built as a multi-layer system deployed on Amazon AWS. It includes an admin web application for managing event sources and scraping schedules, automated data collection services, an AI-powered processing layer, a centralized database, and a public-facing web application for browsing events.
AI is used to process large volumes of unstructured text by generating meaningful summaries, enriching missing descriptions, assigning tags, and filtering out duplicate or irrelevant events.
Built by a five-member student team with guidance from academic and industry mentors.
Students: Kristián Kandra, Kamil Škurla, Matej Maňkoš, Štefan Jaklovský, Margaréta Renčková
Company mentors: Adam Antal (IBM), Peter Butkovič (IBM). University mentor: Sergej Chodarev (KPI)