Photo courtesy of Vara Imandi
Opinions expressed by contributors are their own.
It was a Tuesday afternoon when Maria Lopez, a nurse practitioner at a community clinic in Phoenix, received an alert she had never seen before. A patient with severe asthma needed prior authorization for a lifesaving biologic drug—a process that typically took days of phone calls, faxes, and bureaucratic limbo.
This time, an AI agent intervened. Within minutes, it crossreferenced the patient’s insurance data, clinical history, and treatment guidelines, then autogenerated a complianceready submission. “It felt like magic,” Lopez said. “But really, it’s just the future landing ahead of schedule.”
The “magic” Lopez described is the work of technology innovators like Vara Imandi, a Salesforce executive whose twodecade journey through software engineering, system architecture, and AI deployment has positioned him at the vanguard of healthcare’s quiet transformation.
Hospitals and insurers urgently need to address workforce shortages and rising costs. Meanwhile, Imandi uses AIdriven automation—especially via MuleSoft and Salesforce’s Agentforce—to transform how care is delivered, paid for, and scaled.
Healthcare’s hidden burden meets its AI match
Prior authorization, a process that can delay critical treatments, is a significant burden on healthcare providers. It involves verifying a patient’s insurance coverage for specific treatments, often requiring multiple phone calls and faxes. Lopez navigated the same process, which is a $93.3 billion annual drain on U.S. healthcare.
According to the American Medical Association’s definitive late 2024 physician survey, medical practices now spend an average of 13 hours per physician each week navigating these insurance approvals, while patients face dangerous delays. These are not benign delays; The same survey found that 93% of physicians report that prior authorization delays care, and 29% report that such a delay has led to a serious adverse event for patients, including death. This administrative burden is paradoxically driving up costs and creating the very patient risks it was meant to prevent. Imandi’s solution? AI agents that automate the grind.
“The next wave of healthcare transformation will not be defined by moonshot cures, but by the systemic elimination of operational friction. This is being achieved by invisible, autonomous AI that automates administrative burdens to restore the industry’s humancentric purpose. These aren’t chatbots. They are autonomous digital workers, executing complex, multisystem tasks that were previously the sole domain of human administrative staff.” Imandi explains.
His team’s Healthcare Accelerator, a bundle of prebuilt APIs and compliance frameworks, allows partners like Deloitte, Accenture, PWC, and many others to deploy AI agents that interface with electronic health records (EHRs), insurance portals, and regulatory databases. The result: In pilot programs conducted with earlyadopter health systems, these agents reduced prior authorization timelines by up to 70%, with accuracy rates exceeding 95%.
The impact extends beyond efficiency. In Life Sciences, Imandi’s agents now automate clinical trial scheduling – a task so complex that 80% of trials historically face delays. By syncing participant availability, trial protocols, and site logistics, these tools have reduced noshow rates by 40% in early adopters like Johns Hopkins. “We’re not just saving time,” says a pharmaceutical executive who requested anonymity. “We’re saving science.”
From code to care
Imandi’s path to healthcare AI began 6,000 miles from Silicon Valley. Born in India and trained as a computer engineer, he cut his teeth at Oracle, Dell, and Deloitte, where he designed integration tools for banking and retail. But a pivot to healthcare in 2018triggered by his grandfather’s cardiac emergency, sharpened his focus. “I saw how fragmented the system was,” he recalls. “Doctors had the tools to heal, but not the tools to connect.”
That insight led him to Salesforce, where he now bridges two worlds: the technical arcana of API architecture and the visceral urgency of patient care. His signature achievement—MuleSoft’s Industry Accelerators for Healthcare and Life Sciences—provides partners with prepackaged solutions for healthcare’s thorniest challenges, from priorauth to pharmacovigilance reporting to fraud detection.
Consider Pharmacovigilance: When adverse drug reactions occur, pharmaceutical firms must report them to global regulators within 15 days. Imandi’s AI agents parse EHRs, scientific literature, and patient interactions data to flag potential cases, then autogenerate submissions compliant with the FDA, EMA, and 30 other agencies. For one Top 10 pharma company, this reduced reporting errors by 62% and accelerated submissions by 50%.
The AI ecosystem takes shape
Imandi’s work thrives on a paradox: To make AI transformative, it must become invisible. At Salesforce’s TDX 2025 conference, he demonstrated how the Agentforce generative AI platform orchestrates multisystem workflows. In a live demo, an AI agent diagnosed a billing discrepancy, negotiated with an insurer’s chatbot, and updated a patient’s EHR – all without human intervention.
The real magic, though, happens behind the scenes. Imandi’s team runs a global partner network, training firms like PwC and Accenture to deploy these tools. Through workshops and hackathons, they’ve nurtured a community of 15,000+ developers building AI agents for niche use cases, from prior authorization in rural India to clinical trial recruitment in Lagos. “This isn’t about replacing humans,” Imandi insists. “It’s about letting them focus on what only humans can do.”
The numbers highlight his point: Hospitals using AI agents report 30% faster claim processing and 20% reductions in administrative staffing costs. According to a Q1 2025 forecast from Forrester Research, generative AI adoption in healthcare operations is set to triple by 2027, with a specific focus on automating revenue cycle management and provider credentialing. Forrester projects the market for these specialized AI agents will exceed $12 billion by 2028.
Global implications and ethical frontiers
The ripple effects are global. In the US, Blue Shield of California, an independent member of the Blue Shield Association, founded in 1939 with 6 million members, over 7,500 employees, and more than $27 billion in annual revenue, use Imandi’s frameworks to automate priorauthorization requests for 500,000+ lowincome governmentinsured patients annually deployed on Azure services. In Sweden, AI agents are now being used to coordinate dementia care across 40 clinics, reducing caregiver burnout by 25%. However, challenges persist regarding data bias and transparency.
AI alone cannot solve systemic inequities—biased data will inevitably lead to biased outcomes. Recognizing this risk, Imandi emphasizes the importance of implementing guardrails. His teams routinely audit AI outputs to identify disparities and collaborate with partners to ensure these technologies serve underserved communities equitably.
Still, ethical dilemmas loom. Should an AI agent prioritize a patient’s insurance status when recommending treatments? Who is liable if a prior authorization error causes harm? Imandi’s answer: Transparency. “Every decision must be explainable, every workflow auditable.”
The humanmachine balance
Back in Phoenix, Maria Lopez now spends her freedup time on what matters: facetoface care. “I finally have room to breathe,” she says. That’s the promise Imandi envisions – a world where AI handles the bureaucratic “noise,” letting clinicians reclaim medicine’s human core.
But the road ahead is uneven. While Boston and Berlin embrace AI agents, rural hospitals in Mississippi and Mozambique lack the broadband or budgets to join. Imandi’s next mission? Democratizing access. Through opensource APIs and partnerships with nonprofits, he aims to work with more partners to bring these tools to 100+ lowresource regions by 2026.
“Technology alone won’t heal us,” he reflects. “But when you pair it with human compassion, that’s where the future gets interesting.”
The takeaway
The AI breakthrough in healthcare is not a distant promise; it is a presentday reality, operating quietly within the industry’s essential, mundane workflows. The work of innovators like Vara Imandi demonstrates that the most powerful transformations arise from systematically eliminating the operational friction we have long accepted as the cost of doing business.
As healthcare continues to integrate AI, the crucial question still remains: Will these technologies enhance the human touch in care, or diminish it? The answer rests squarely in our hands. After all, agents can automate workflows, but they cannot replicate professional judgment or automate empathy. The ultimate measure of this transformation won’t be found in efficiency gains, but in whether a nurse like Maria Lopez gains the precious moments to hold her patient’s hand, or tragically, loses it.
