According to CDC, more than a million emergency department visits yearly are caused by adverse drug events. This is a massive issue Reckitt pays a lot of attention to, putting the safety of our consumers at the core of everything we do. The company is also looking to extend the monitoring of potential adverse reactions to non-medical products from FMCG sector. This is why the data science team at Reckitt is building a scalable machine-learning system based on a unique proprietary dataset leveraging contemporary NLP techniques to identify and categorize potential adverse events from consumer reviews on e-commerce platforms and social media. Combining text classification and named entity recognition, the system will allow for identifying early signals of product issues online, leveraging the opportunity to react to the problems before the reports from official sources are received.