Over 99% of trips on the Lyft platform occur without any reported safety incidents; even when safety incidents do occur, only 0.0002% of trips are considered severe (i.e., motor vehicle fatalities, fatal physical assaults, certain categories of sexual assault). To proactively help users in the case of a rare incident, Lyft monitors rides by using objective ride factors to detect irregularities and proactively check-in with drivers and riders asking if they need support – and if necessary, request emergency assistance. However, even in instances where we notice a ride that appears to be irregular and could indicate a potential safety incident (this may include rides that have ended far from the intended destination, been canceled after pickup, gone off-route or stopped for an abnormal amount of time), the majority of trips have no reported issues due to benign circumstances, like unexpected roadwork or traffic. So how do we design an experience that proactively connects users to the right safety tools when we detect potentially unsafe situations, given this high uncertainty and propensity to forget about available tools in high stress situations? In this presentation, I will cover the Lyft Safety team’s strategy and learnings for addressing uncertainty in safety experiences with our launch of Smart Trip Check-in. Specifically, it will help the T&S professional community: 1. Develop principles to facilitate better and faster decisions making in an ambiguous space such as safety (example: a key design principle was assuming good intent but preparing for worst case scenario) 2. Evaluate inherent tradeoffs across user segments (e.g. riders vs drivers) and goals to maintain a positive user experience (e.g how to find the sweet spot between false positives and false negatives) 3. Solve for a spectrum of potential unsafe situations, providing the right tools across severity 4. Design flexible systems that allow for quick test and iterations (diving into design and messaging iterations and learnings)