In todayâs healthcare landscape, administrative waste has quietly become one of the industryâs most expensive and least examined problems, a drag on hospitals, clinicians, and patients alike that technology has mostly made faster rather than actually fixed. The Healthcare We Were Promised, a new book by Abhinav Shashank, Co-Founder and CEO of Innovaccer, argues that the fix isnât another point solution bolted onto a broken system. Itâs a fundamentally different architecture for how healthcare operates.
Autonomous Operations: The Core of the Bookâs Thesis
At the center of Shashankâs argument is a simple but uncomfortable claim: healthcareâs administrative burden isnât a side effect of complexity; itâs a design failure. Todayâs systems- EHRs, claims platforms, payer databases- donât speak the same language, so armies of coders, billing staff, and care coordinators exist purely to translate between them. The book makes the case that AI tools that simply make that translation faster donât solve the problem; they just make the same broken process run at higher speed. Instead, Shashank proposes a shift toward âintelligent workflowsâ: closed loops that sense a trigger, pull the right data, decide, and act without a human intermediary, reserving human judgment for the moments that actually require it. He maps this transition across five measurable stages of autonomy, from todayâs fully manual baseline to a fully autonomous Level 4.
A Decade in the Making
Unlike a theoretical thought experiment, the bookâs foundation comes from ten years of real deployment. Shashank founded Innovaccer in 2014 to solve what he believed was healthcareâs core problem: fragmented data. Over the following decade, the company built toward that belief at scale, connecting data for 200 million patient lives across 100 health systems. It was watching how that connected data actually got used- a nurse manually re-triaging an AI-generated list for four hours each morning, a billing team routing AI recommendations through an eight-day human review- that led Shashank to a second, deeper diagnosis: the humans sitting between the data and the decision werenât adding value, they were just absorbing the cost of a system that was never designed to talk to itself.
Addressing Gaps in Healthcareâs Administrative Architecture
The scale of the problem the book documents is substantial. Administrative activity that touches no patient and improves no outcome costs the U.S. healthcare system an estimated $750â935 billion annually which is more than the country spends on K-12 education. The book traces this cost to specific, recurring handoffs: a physician spends roughly 45 minutes documenting a complex clinical encounter, only for a coder to spend another 20â30 minutes translating that same documentation into billing codes, introducing error and delay at every step. Provider-to-payer friction such as, prior authorizations, appeals, and claims, adds another $54 billion a year, while poorly managed discharges cost Medicare an additional $26 billion annually in preventable readmissions. Shashank frames the opportunity on the other side of fixing this architecture at a conservative $1 trillion in industry-wide annual savings by 2036, driven by a compounding advantage: organizations that automate first learn faster, widening the gap over every health system that waits.
Validation from Health System and Clinical Leaders
The bookâs argument is reinforced throughout by on-the-record perspectives from leaders inside the organizations actually running healthcare today, lending it a credibility a purely theoretical framework wouldnât have. Brandon J. Burket, Vice President of Value-Based Care and Population Health at Orlando Health, cautions that without workflow redesign, âyouâve just made the existing broken process faster... thatâs more noise, not less, that every stakeholder in the chain has to absorb.â Peter Pronovost, MD, PhD, FCCM, Chief Quality and Clinical Transformation Officer at University Hospitals Cleveland, adds that âAI is an extraordinary amplifier of human capability, but it cannot compensate for weak leadership, fragmented management systems, or cultures that learn slowly.â And Vishal Agrawal, MD, CEO of Longitude Health, grounds the stakes in people rather than process: âtransformation fails when the people who do the work arenât the ones redesigning it.â
From Theory to Practice: A Morning in 2030
Rather than leave autonomy as an abstraction, the bookâs closing chapter walks through a single, technically-grounded morning inside a fictionalized health system before a single human administrator has sat down at their desk. In that window, an autonomous system catches a chemotherapy dose calculated 18% too high before the patient arrives for infusion, correctly triages a 2:47 AM call about chest pain as a cardiac emergency, monitors sepsis risk across 1,247 inpatients overnight, and identifies a nursing staffing gap 72 hours before it would have become a patient-safety issue. Shashank is careful to note that every capability described is already deployed somewhere today. The only fiction is that no single health system has yet stitched them all together into one platform.
The Future of Autonomous Healthcare Operations
With a decade of deployment experience behind it and a trillion-dollar opportunity ahead, The Healthcare We Were Promised positions itself less as a forecast and more as a blueprint, one aimed at executives, technologists, and health system operators âwho are done waiting for someone else to define what healthcareâs AI transformation actually means.â As Shashank puts it, the goal isnât to replace the people delivering care, but to remove the translation layer standing between them and the patients theyâre trying to help, freeing clinicians and administrators alike to spend their time on the work that actually required a human being in the first place.