Tuberculosis (TB) has been with humanity for a very long time and is one of the oldest diseases known to affect humans and top cause of infectious death second to COVID-19. Advancement in diagnostic modalities posts a significant improvement in diagnostic arsenal, but the cost is relatively high compared to simple bacilli detection with microscope. Tuberculosis unfortunately is prevalent in moderate to poor resources countries making the ideal choice for diagnosing TB a cheap, manual process of bacilli detection using microscope. This process requires experienced grader for two grader systems. Presently, we have inadequate personnel to play these roles. Various efforts to automate this process using artificial intelligence (AI) range from adaptive color thresholding, shape descriptors, or image detection via different color spaces. Despite various modalities of artificial intelligence products exist, challenges in adoption to clinical practice persist. Some of the challenges are due to inability to achieve seamless integration with existing workflow, lack of trust, transparency, and acceptance in AI products. Early involvement of clinical teams in development of AI may improve this situation; one way of inciting the interest of clinical teams is by streamlining the development of AI with clinical workflow. Thus this study will suggest a framework for clinical teams to participate in AI utilization via introduction of digitization process aligned with existing clinical workflow for microscopy analysis of tuberculosis smear samples stained with fluorescein.