OptySleep empowers users with personalized sleep tracking and insights via OptyAI, a Sleep Widget Tracker, and quick data logging. The OptySleep Engine—a recommendation engine built from scratch with a custom formula, leveraging Redis caching, A/B testing, FastAPI, and Docker—was designed in-house to address performance issues, resulting in a ~70% boost in engagement, improved retention, and higher subscription renewals.
The system uses natural language processing and semantic similarity techniques (Sentence-BERT embeddings) to predict next sleep methods based on previous user trials. The custom formula analyzes sleep histories to provide personalized recommendations, such as suggesting 'blue light filters' if 'caffeine reduction' improved energy levels.