AMA: Asesso Health Chief Product Officer, Abhijit Limaye on Roadmapping /Vision Setting and the use of AI in the Health Tech Industry

In Health tech, agility isn't just a buzzword—it's a survival mandate. You will need to be nimble, and continuously evaluate how (or if) recent technological inflections are shifting your foundational product strategy. Specifically, have these advancements materially changed your Ideal Customer Profile (ICP), OR to what extent can they be leveraged to widen your competitive moat? In my experience, if you have the right North Star vision, you have a great product-to-market fit, it matches your co ...Read More

We need to distinguish between AI as a 'visionary' concept and AI as a 'validated' tool. The hype usually lives in the vague promise of 'what AI could do.' The reality is that we are still in the early stages of building models with the deep, domain-specific context required for healthcare. To lead effectively in this space, we have to move past the marketing-hype behind AI. Focus on these 3 things when you think about how to use AI in your workspace. 1. The 'Co-pilot' vs. 'Autopilot' Distinctio ...Read More

Alignment is not about compromise; but rather about calibration and a common understanding. Each team is working to maximize their currencies of success": Clinical (safety/evidence), Engineering (velocity/scale), and Commercial (ROI/market share), and the risk is that each operates in a silo. So recommending a framework to bring the different silos together. 1. Define the "Unified North Star" Anchor all decisions in some patient/customer centric. When priorities clash, this North Star serves as ...Read More

I’m mostly in a 'watch and learn' mode, focusing on the structural shifts that actually move the needle rather than chasing every shiny object. The FDA vs. CMS Gap The biggest trap I see right now is the disconnect between getting cleared and getting paid. The FDA validates that your tool is safe, but CMS is the one that decides if it’s actually useful. Focus on building something that is 'reimbursement-ready' evidence from day one so we don't end up in that limbo. I’m also keeping a close eye o ...Read More

In health tech, confidence is not won through enthusiasm; it is won through a rigorous framework of validation and risk management. Don't treat AI as a monolith - "Technology" is too broad a term for an executive roadmap. To build trust, you must categorize your AI initiatives by their maturity level. For example, traditional machine learning (classification, regression) is already stable and reliable for predictive analytics. Generative AI and LLMs require more scrutiny. Design visible guardrai ...Read More

We are making 2 big transformational bets actually, but both are long term bets for the industry and our company. As such either is not in the immediate critical path for commercialization but sufficiently near-to-mid term to cause disrupt. Clinical Diagnosis - We are betting that FDA will become more open using AI for diagnosis, as long as the prognosis is made based ample evidence-based systems and guardrails. Developing a foundational visual learning model for dentistry - Building such models ...Read More

This depends largely on your team’s current maturity and bandwidth. You may want to develop these skills further but understanding what you have today, is important so that you can separate real utility from noise. Here are some tips / thoughts - 1. Core Competency - Build only what sits at the heart of your intellectual property and competitive advantage and adds to the intrinsic value of your company. Buy for operational efficiency and mission-critical stability. If a proven solution exists fo ...Read More

This is a bit contextual since Healthcare is a vast field. Augmenting clinicians - Be sure to distinguish between AI as a 'visionary' concept and AI as a 'validated' tool. The 'Co-pilot' vs. 'Autopilot' Distinction Today, AI’s highest and best use is as a clinical decision support tool. It is an enabler that helps clinicians surface insights and speed up workflows. So use AI as a 'Force Multiplier, keep the human-in-the-loop, which mitigates clinical risk and eases the regulatory burden. Replaci ...Read More