The a16z Show Episode: Ambience CEO Nikhil Buduma on AI in Clinical Workflows Host: Andreessen Horowitz (Julie Yu) Guest: Nikhil Buduma (CEO & Co-Founder, Ambience Healthcare) Date: March 4, 2026
Episode Overview
This episode dives deep into how artificial intelligence (AI) is transforming clinical workflows. Julie Yu (a16z General Partner) interviews Nikhil Buduma, CEO and co-founder of Ambience Healthcare, about the evolving landscape of AI in healthcare, the technological and operational challenges, the promise of automation, and the unique journey behind founding Ambience. The conversation oscillates between clinical realities, organizational adaptation, and the “lava floor” pace of AI and product evolution.
Key Discussion Points & Insights
1. The State of Healthcare and AI's Opportunity
- The demand for healthcare is outpacing physician supply, with over 10,000 people aging into Medicare daily ([00:00], [28:23]). AI offers a path to "do more with less."
- There’s growing hope that AI can restore joy to the clinician experience and reduce patient confusion ([00:39], [44:32]).
Quote:
"There is a pathway for the job of being a clinician, being a nurse, to be a fulfilling one. There is a pathway for the experience of a patient not being as confusing and full of despair as sometimes it is."
— Nikhil Buduma [00:39], repeated [44:32]
2. Nikhil Buduma’s AI & Healthcare Journey
- Originally aimed to be an MD/PhD before shifting to tech after witnessing a mentor’s death due to a medical error ([02:32]).
- Early involvement with deep learning research, OpenAI’s formative group, and observing the “transformer” paradigm shift ([02:32] – [05:14]).
- Founded Ambience after running a medical practice firsthand to deeply understand workflow pain points ([02:32] – [05:14]).
- Ambience now works with large academic medical centers, achieving daily use by 75%+ clinicians and driving significant financial impact ([00:56], [12:58], [35:36]).
Quote:
"We started a care delivery asset and...were taking all of the techniques that we were seeing in the research labs and bringing them into the practice...that experience became the foundation for Ambience."
— Nikhil Buduma [02:32]
3. Product & Market Evolution: Infrastructure and Segmentation
- Early AI required custom training, infrastructure, and long iteration cycles ([05:45]):
“The iteration cycles were insane...iteration cycles were like a year.” — Nikhil Buduma [05:45] - Ambience began as a full-stack care provider, then pivoted to building a platform after developing empathy for on-the-ground users ([06:58]).
- Market segmentation:
- Large, complex health systems (Ambience’s forte): higher adoption, tougher for generalist AI.
- Smaller practices: lower complexity, more competition and commoditization ([09:11] – [12:58]).
- Massive adoption challenge: many competitors fail to get more than 20% utilization.
Quote:
"There's a high complexity part of the market that's really, really hard to serve. But if you can serve it, it's hard for others to compete in."
— Nikhil Buduma [12:58]
4. AI’s Evolving Role: From Intelligence Layer to “Action” Layer
- Debate on whether general-purpose AI will commoditize the “clinical intelligence” layer, shifting value to agents that take action ([13:34] – [17:49]).
- Nikhil argues the “floor is lava”—AI capabilities race ahead, but last-mile product, data, and workflow problems remain long-term moats ([14:47]).
- Persistent data challenges: integrating messy, inconsistent EHR data, tracing clinical decisions, defining quality ([16:35] – [18:38]).
- Deep, fast iteration with customer partners is crucial—Ambience can rapidly move from prototype to deployment within 30 days ([18:38] – [20:48]).
Quotes:
"AI clock speed is fundamentally different from product clock speed...building a world where the floor is lava."
— Nikhil Buduma [14:47]
"Defining quality is actually just a fundamentally hard problem and one where it has to be solved at the intelligence layer…"
— Nikhil Buduma [17:49]
5. The Battle for the EHR & Backend Infrastructure
- The possibility of disrupting incumbent EHRs by introducing AI-native layers ([24:48] – [26:38]).
- Ambience’s innovation: building a data layer above the EHR that normalizes input for AI-driven applications, vastly accelerating product deployment ([24:48] – [26:29]).
- This infrastructure shifts “product clock speed,” unlocking dozens of new features and apps rapidly ([26:31]).
Quote:
"Once you've had that infrastructure layer that we've created, it fundamentally changes your product clock speed as an organization."
— Nikhil Buduma [26:29]
6. Toward Autonomous AI and the “Virtual Care Team”
- Current phase: AI as “copilot”—clinicians still sign off on documentation ([27:51]).
- Next phase: Ambience working on anticipatory summaries, before-visit prep, and post-visit patient engagement agents ([29:33]).
- Vision: virtual care team members taking over growing swathes of clinician workload—continuous patient conversations, proactive care support ([29:33] – [30:41]).
Quote:
"Imagine a clone of yourself...has pored through all the data for hours and hours and hours and has put together a summary...What does it look like for now there to be a virtual care team member to actually help quarterback all of those things on your behalf?"
— Nikhil Buduma [29:33]
7. Practical Advice for Health System Leaders
- Two key questions for buyers ([31:35] – [35:36]):
- Adoption—Will your clinicians actually use it? Utilization is the leading indicator of value.
- Operating Margin Impact—Does the AI platform drive real, attributable financial results?
- The “flywheel” effect: AI→ROI→investment in talent→higher volume→further AI adoption ([32:17], [35:36]).
Quotes:
"There's a big difference between marketing and noise versus if I bought this, will my clinicians actually use it?"
— Nikhil Buduma [31:35]
"One center we work with, they're projecting over $30 million of net new margin...just created by ambience because they don't have to hire human scribes anymore."
— Nikhil Buduma [35:36]
8. The Provider–Payer AI Arms Race
- As both payers and providers deploy AI for revenue cycle management (RCM) and payment integrity, ultimate industry benefit lies in standardized, high-fidelity sources of truth ([37:34] – [39:19]).
- If both sides achieve data integrity, incentives to “fight” reduce—collaboration likelier, and administrative costs could fall ([37:59]).
Quote:
"Once you have a shared source of truth, then the ROI of RCM becomes potentially negative over the next five years...it just makes sense to collaborate."
— Nikhil Buduma [37:59]
9. Remaining Challenges and Future Vision
- Still hard: cascading context across complex care settings; some predictive modeling ([39:59]).
- The main bottleneck now is not model capability but the “last mile” — workflow integration and clinical/RCM subject matter embedding ([40:00] – [41:14]).
- Platform future: Ambience contemplating becoming a backend layer for third-party devs ([41:14] – [41:56]).
- Internal operations: Dramatic productivity gains from AI tools—fewer people, smarter thinkers, deeper subject matter expertise valued ([42:19] – [42:55]).
Memorable Moments & Quotes
-
On Product Adoption:
“You can literally just let the doctors use it and kind of put with their own feet, right?”
— Julie Yu [32:17] -
On Clinical Joy:
“I've never experienced the type of joy that I've experienced using this tool in my day to day job that has made me now want to remain a doctor.”
— Doctor to Julie Yu [45:52] -
On Cultural Shift:
“The delta between the magic of the tools that they're experiencing in their consumer lives and what they do in their work has for the first time narrowed...fundamentally changed the nature of how they view technology.”
— Julie Yu [00:20], echoed at [45:52] -
On Building in Healthcare:
“There is a specific level of just humility and honor in being able to work on this problem...it’s just one of those industries where the outlook looked bleak...this is the first time where there's hope.”
— Nikhil Buduma [44:32]
Timestamps for Important Segments
- 00:00 – The need for change: healthcare demand, workforce shortage, AI hope
- 02:32 – Nikhil Buduma’s journey, OpenAI origins, founding Ambience
- 05:45 – Early AI technical challenges and iteration cycles in healthcare
- 06:58 – Full-stack to platform transition: Remedy to Ambience
- 09:11 – Provider market segmentation and AI adoption patterns
- 12:58 – The “complexity gap” and why generalized AI struggles in healthcare
- 14:47 – Intelligence vs. action layer; “floor is lava” dynamics
- 16:35 – Data integration, decision trace, and quality definition challenges
- 24:48 – Disrupting the EHR and backend stack
- 27:51 – Autonomous AI doctor: barriers & the shift from copilot to agentic systems
- 29:33 – Pre- and post-visit AI: the “virtual care team” vision
- 31:35 – Advice for health system CIOs: prioritize adoption and ROI
- 35:36 – Real financial results, the AI flywheel, and shifting ROI focus
- 37:34 – Payer-provider AI arms race: can source-of-truth lead to collaboration?
- 39:59 – Remaining technical/product bottlenecks
- 41:14 – Ambience’s possible evolution into an open platform
- 42:19 – Internal transformation: AI-native company operations and hiring
- 44:32 – Personal motivations, healthcare's new sense of hope
- 45:52 – Clinician testimonials: from burnout to joy
- 47:21 – The “iPhone effect” for doctors and tech’s new reputation in healthcare
Final Thoughts & Tone
Throughout the conversation, both speakers remain practical, optimistic, and candid about limitations, culture shifts, and the deep responsibility that comes with building transformative tools for healthcare.
Closing reflection:
"I think this is the first time where there's hope that hey, there is a pathway to doing more with less…And I think in some ways that makes this a special time to be working in healthcare."
— Nikhil Buduma [44:32]
This episode is essential listening for healthcare executives, technologists, and anyone interested in the convergence of AI and real-world clinical impact.