Understanding how stem cells develop and differentiate is vital for research and biomanufacturing, but existing methods can’t track this process in real-time. To address this, researchers developed a technique using fluorescence lifetime imaging (FLIM) with machine learning to monitor stem cell changes as they happen. They identified 56 key markers that reveal shifts in hematopoietic stem cell (HSC) development, including choices in their growth paths. The method also created a “metabolic stemness” score to assess stem cell quality, providing insights into how to improve stem cell maintenance and production outside the body.