OpenScout: An Autonomous Open-Weight AI Event Scout

OpenScout is an autonomous AI-powered event notifier designed to help developers stay informed about relevant conferences, meetups, and hackathons that align with their specific interests. The system was created for a friend who is passionate about open-source development, AI/ML, DevOps, Docker, Kubernetes, and Cloud Computing. This friend often missed out on valuable events due to the scattered nature of event announcements across various websites, RSS feeds, Discord communities, and social media.

To address this issue, the developer built OpenScout, which aggregates open-source opportunities, understands the user's technical interests, ranks matches using AI, and alerts them when an event is worth attending. OpenScout utilizes open-weight AI models, specifically Google's Gemma (gemma2:2b / gemma3:4b), which are run locally via Ollama. This approach ensures complete data privacy and sovereignty, eliminates recurring token costs, and provides full transparency and customizability.

The system is composed of several components: a FastAPI backend with clean REST endpoints and an extensible EventSource architecture; an open-weight AI relevance engine built for local inference; a SQLite persistence layer for profile preferences, event storage, user feedback, and notification logs; a frontend dashboard built with accessible, responsive HTML5/CSS/Vanilla JavaScript; and an autonomous background runner with Docker configurations.

The open-weight model processes event details and outputs structured JSON containing a relevance score, matched interests, a human-readable explanation of why the event fits, and a notification flag. User feedback is stored in SQLite to build a personalized dataset for future few-shot prompt adaptation.