Though the potential of artificial intelligence (AI) in healthcare warrants genuine enthusiasm, the road to meaningful impact is long and fraught with challenges. AI tools are susceptible to mistakes and rarely capable of capturing all of the nuances pertaining to a complex clinical situation. However, these challenges are not insurmountable. In this talk, I will first highlight our work in overcoming technical barriers related to model robustness. Specifically, I will describe novel approaches for improving model generalizability through the incorporation of domain expertise. In the second part of the talk, I will describe our efforts in overcoming implementation barriers by engaging stakeholders and carefully considering existing clinical workflows. In summary, there’s a critical need for AI in healthcare; however, the safe and meaningful adoption of these techniques require extensive interdisciplinary collaboration.