Artificial intelligence has become one of healthcare’s most heavily discussed technologies, but investor Kam Thindal believes much of the conversation misses where adoption will actually happen first.
Thindal, Managing Director of Core Capital Partners, argues that healthcare’s greatest inefficiencies are operational rather than clinical. Instead of replacing doctors, he expects AI to reduce the administrative burden surrounding patient care.
That burden remains significant across hospitals and clinics. Healthcare workers continue to manage extensive paperwork, insurance approvals, documentation requirements, coding systems, and disconnected digital platforms.
“If you want to predict where AI penetrates first, follow the pain, not the headlines,” Thindal says.
He believes the strongest early use cases involve repetitive, high-volume tasks such as prior authorization workflows, chart summarization, coding support, scheduling, and claims management.
“A model that drafts a prior authorization appeal, summarizes a chart, or flags missing coding elements can create real leverage,” he explains.
Thindal says the timing for healthcare AI adoption is being shaped by several overlapping pressures. Clinician burnout continues to strain staffing capacity, while healthcare systems now generate far more digital data than they did a decade ago. At the same time, AI models have become better at interpreting conversations, forms, notes, and other unstructured information.
“The leap from narrow automation to language and multimodal systems expands what can be automated,” he says.
Still, he warns investors against assuming healthcare will adopt AI at the pace seen in consumer technology. Unlike social media or productivity apps, healthcare organizations operate under strict regulatory oversight and face serious liability risks.
“A hallucination in a medication list is dangerous,” Thindal says.
Because of that, he expects healthcare AI adoption to happen gradually. Administrative tools will likely expand first, followed by documentation support and constrained clinical decision systems where outputs remain easier to verify.
He also believes integration will determine which companies succeed. AI products that sit outside existing workflows may struggle to gain traction inside busy clinical environments.
“Tools that live outside the workflow die in the workflow,” he warns.
For Thindal, the long-term opportunity remains substantial, but investors must approach the sector with realistic expectations about adoption timelines.
“Healthcare will not be disrupted in one sweep,” he says. “It will be rewired module by module, and trust will be earned through consistency over time.”