Around 3.5 million news articles are published on the public internet every day, making it impossible for any individual to read them all. The Noisefilter began with a simple question: what if these articles could be grouped into stories and used to show how news spreads across the world?

The platform was designed for decision-makers, consulting teams, and diplomatic delegations that previously relied either on a handful of major news outlets, such as the BBC, Financial Times, The Guardian, and Reuters, or on dedicated communications teams preparing daily media briefings.

LogiNet developed the news intelligence platform that monitors articles from around the world, groups them into related stories, and presents the results through The Noisefilter's own analytics in a single interface. Operating 24/7, the platform processes up to 60,000 articles every day from trusted external news sources in real time. A Kotlin-based microservices backend runs every incoming article through an AI pipeline, while an interactive SVG-based world map visualises how each story spreads across countries.

Grouping articles into stories is primarily a data science challenge. The process begins with database-level similarity matching using fuzzy and semantic search, combined through rank fusion techniques. An additional LLM layer then identifies hidden relationships and filters out false matches. This allows articles covering the same event from different countries and perspectives to be grouped into a single story.

The platform has been operating in production since mid-2025. It is designed for professionals who want to understand not only what happened, but also how a story develops globally and how the same topic is reported in different regions. The platform places major international events into context while also revealing the global impact of local news. Decision-makers can gain an overview of the world's news landscape in around 15 minutes.