Why AI Could Finally Fix Healthcare Access | Zocdoc CEO Oliver Kharraz
Episode
34 min
Read time
2 min
Topics
Productivity, Health & Wellness, Relationships
AI-Generated Summary
Key Takeaways
- ✓Provider Directory Accuracy: Static provider directories fail because they function like Encyclopedia Britannica — outdated the moment published. ZocDoc achieved 99%+ data accuracy by making directories transaction-based, like Wikipedia, where each completed booking validates and corrects provider information in real time, eliminating the perpetual data-chasing problem that plagues insurance directories.
- ✓EHR Integration Depth: Connecting to an EHR is straightforward; making scheduling truly accurate requires solving roughly 100 edge cases per system. ZocDoc discovered that office managers routinely used Post-it notes instead of updating calendars, causing booking errors. Identifying and mathematically accounting for these human workarounds across 175+ EHR systems took nearly two years.
- ✓Last-Minute Capacity Recovery: National average wait times for primary care have risen from 17 to 31 days since ZocDoc launched, yet ZocDoc users book appointments within 3 days on average. The mechanism is last-minute cancellation recovery — similar to how Expedia and Kayak lifted airline load factors from 60% to 90%, ZocDoc raises physician utilization from roughly 70% toward capacity.
- ✓AI Agent Infrastructure Strategy: Rather than defending ZocDoc.com as the single booking destination, ZocDoc opened its scheduling API to AI agents, payers, and third-party platforms. Providers complete one integration; ZocDoc distributes their real-time availability everywhere. This positions ZocDoc as healthcare scheduling middleware — capturing booking volume regardless of which AI agent or platform the patient uses.
- ✓Consumer Loyalty Data: Locking patients into a single health system through restrictive scheduling is unnecessary. ZocDoc data shows a 75–80% rebooking rate with the same provider within the same specialty after an initial booking. Patients self-select loyalty when they receive good care, meaning open scheduling increases provider utilization without sacrificing patient retention.
What It Covers
ZocDoc CEO Oliver Kharraz explains how the company built scheduling infrastructure connecting patients to doctors across 175+ EHR systems, why national appointment wait times have doubled to 31 days since ZocDoc launched, and how AI agents and payer partnerships like Blue Shield of California are expanding healthcare access.
Key Questions Answered
- •Provider Directory Accuracy: Static provider directories fail because they function like Encyclopedia Britannica — outdated the moment published. ZocDoc achieved 99%+ data accuracy by making directories transaction-based, like Wikipedia, where each completed booking validates and corrects provider information in real time, eliminating the perpetual data-chasing problem that plagues insurance directories.
- •EHR Integration Depth: Connecting to an EHR is straightforward; making scheduling truly accurate requires solving roughly 100 edge cases per system. ZocDoc discovered that office managers routinely used Post-it notes instead of updating calendars, causing booking errors. Identifying and mathematically accounting for these human workarounds across 175+ EHR systems took nearly two years.
- •Last-Minute Capacity Recovery: National average wait times for primary care have risen from 17 to 31 days since ZocDoc launched, yet ZocDoc users book appointments within 3 days on average. The mechanism is last-minute cancellation recovery — similar to how Expedia and Kayak lifted airline load factors from 60% to 90%, ZocDoc raises physician utilization from roughly 70% toward capacity.
- •AI Agent Infrastructure Strategy: Rather than defending ZocDoc.com as the single booking destination, ZocDoc opened its scheduling API to AI agents, payers, and third-party platforms. Providers complete one integration; ZocDoc distributes their real-time availability everywhere. This positions ZocDoc as healthcare scheduling middleware — capturing booking volume regardless of which AI agent or platform the patient uses.
- •Consumer Loyalty Data: Locking patients into a single health system through restrictive scheduling is unnecessary. ZocDoc data shows a 75–80% rebooking rate with the same provider within the same specialty after an initial booking. Patients self-select loyalty when they receive good care, meaning open scheduling increases provider utilization without sacrificing patient retention.
Notable Moment
Kharraz describes a New York academic medical center dean who wanted to leave ZocDoc because he was booked three months out. After staying 30 more days, roughly 40 patients filled his last-minute cancellations through ZocDoc — slots that would have gone completely unused without the platform.
Episode Transcript
In travel, you book a flight. In ecommerce, you buy a product. But in healthcare, simply booking an appointment can still take weeks. In this episode, a sixteen z general partner Julie Yu sits down with ZocDoc founder and CEO Oliver Caraz to discuss why healthcare scheduling remains so difficult and how AI could change that. Oliver explains how ZocDoc built the infrastructure to connect patients with doctors across hundreds of scheduling systems and why the company is now making that technology available to AI agents in other platforms. They also discussed Sockdoc's partnership with Gemini, why no single company should own the front door to health care, and how better scheduling can help patients see doctors in days rather than weeks. And ultimately, what changes when finding and booking the right doctor becomes as easy as booking a flight? I am so excited to be here with Oliver Caraz, who is the CEO of ZocDoc. Oliver and I go way back. You and I are kindred spirits in both having tried to solve a lot of similar problems in health care over the years, and dare I say, decades almost. And I feel like we competed at times. We probably should have just partnered with each other, but it's been fun to share notes with you as DocDoc has continued to evolve. And, arguably, DocDoc is one of the few companies in health tech that has truly built, like, a household iconic name for itself. So huge kudos for doing that. Maybe Oliver, we can start with you just providing the audience with an update. What is ZocDoc today, and how has it evolved the most over the last several years that you've been building the business? Yeah. Thank you, Julie. Good to be here. And in fact, it's been decades, but it feels like days, and I always enjoyed learning from you. So a lot of your listeners will probably think of ZocDoc as a destination website where they go and can find a doctor that accepts their insurance, and then they'll book an appointment right there and then. And that's something that we, of course, still do. But what is new now is that we take this infrastructure that we have built that enables that marketplace, and we make it available to third parties. So whether you are a payer, you're an AI platform, you have a consumer audience, you can now give them the ability to schedule their doctors seamlessly no matter where they start the journey. The genesis of this conversation actually was you and I just texting each other after I put a tweet out there a few weeks ago saying tongue in cheek that nothing is certain in life except death, taxes, and the provider directory problem. And you and I have both tried to solve this problem from different angles, but it still persists. Right? I mean, that's part of why I wrote it was we continue to encounter, whether it's our …
Get the full transcript (6,263 words) + summary by email — free
One-time email with the complete transcript and AI summary of this episode. No account needed.
One email, no spam. We’ll also show you what SignalCast does.
You just read a 3-minute summary of a 31-minute episode.
Get a16z Podcast summarized like this every Monday — plus up to 2 more podcasts, free.
Pick Your Podcasts — FreeKeep Reading
More from a16z Podcast
AI Can Write the Code. Can It Run the System? | Grepr CEO Jad Naous
Oct 16 · 60 min
Latent Space
Why AI Infrastructure must evolve for Agent Experience — Akshat Bubna, Modal CTO
Jul 8
More from a16z Podcast
When AI Agents Start Spending Your Money | Alex Rampell & Terry Angelos
Oct 15 · 51 min
Invest Like the Best with Patrick O'Shaughnessy
William Hockey - Building the Operating System for the Dollar and Silicon Valley Heresy - [Invest Like the Best, EP.463]
Mar 17
More from a16z Podcast
We summarize every new episode. Want them in your inbox?
AI Can Write the Code. Can It Run the System? | Grepr CEO Jad Naous
When AI Agents Start Spending Your Money | Alex Rampell & Terry Angelos
Can Open Models Keep Up With the AI Frontier? | Nathan Lambert
Can AI Break Cryptography? What the Experts Actually Think
The Race to Recursive Self-Improving AI
Similar Episodes
Related episodes from other podcasts
Latent Space
Jul 8
Why AI Infrastructure must evolve for Agent Experience — Akshat Bubna, Modal CTO
Invest Like the Best with Patrick O'Shaughnessy
Mar 17
William Hockey - Building the Operating System for the Dollar and Silicon Valley Heresy - [Invest Like the Best, EP.463]
Business Of Biotech
Jan 12
A Royalty Model For Value Creation With Zymeworks' Kenneth Galbraith
Invest Like the Best with Patrick O'Shaughnessy
Oct 13
David Vélez - Building Nubank and the Future of Global Banking - [Invest Like the Best, EP.495]
All-In with Chamath, Jason, Sacks & Friedberg
Sep 20
Adam Foroughi, Applovin CEO: Surviving a 92% Drawdown, Ads as ML 1.0 & the $50B Game Ad Market
Explore Related Topics
This podcast is featured in Best Business Podcasts (2026) — ranked and reviewed with AI summaries.
Read this week's Health & Longevity Podcast Insights — cross-podcast analysis updated weekly.
You're clearly into a16z Podcast.
Every Monday, we deliver AI summaries of the latest episodes from a16z Podcast and 192+ other podcasts. Free for one show.
Start My Monday DigestNo credit card · Unsubscribe anytime