AI-powered analysis of social media and web data to detect early signals of market misconduct.
We partnered with the SBS of Peru to design and deliver a working prototype that leverages artificial intelligence and advanced analytics for market conduct supervision.
The project aimed to overcome challenges in processing vast amounts of unstructured, web-based data related to financial services, enabling SBS to move from reactive to proactive oversight. Key steps included:
• Diagnostic and design: Identified pain points in existing tools, mapped supervisory needs, and defined requirements for automated monitoring and analysis of online public data.
• Lean procurement: Ran a global competitive process to select a technology vendor capable of developing advanced natural language processing (NLP) and machine learning models for this use case.
• Agile development: Collaborated with Financial Network Analytics and Winnow Technologies to create a prototype capable of scraping and analysing hundreds of thousands of public posts, classifying them by topic and sentiment, and providing real-time alerts.
• Interactive dashboards: Delivered data visualisations that allow SBS supervisors to track consumer sentiment, detect anomalies, and flag potential misconduct across financial institutions more effectively.
This solution demonstrates how AI and data science can expand supervisory capabilities, allowing financial authorities to spot risks earlier, investigate faster, and make more informed policy decisions. The approach avoids vendor lock-in, granting SBS a perpetual license and flexibility to refine and scale the solution in-house or with other partners.
