Podcast Summary: Using AI at Work

Episode 83: Using AI to Scale Marketing and Revenue Teams with Patrick Leung
Date: December 22, 2025
Host: Chris Daigle
Guest: Patrick Leung, CTO of Faro Health

Overview

This episode explores how AI—especially generative AI—is transforming the design and execution of clinical trials, with a focus on Faro Health’s work in dramatically reducing cost and time-to-market for breakthrough medical treatments. Chris Daigle and Patrick Leung dive deep into practical enterprise uses of AI, strategies for recruitment and upskilling, governance, and how AI is impacting both technical and non-technical business functions.

Patrick, drawing on his background at Google Duplex and hedge funds, discusses how Faro Health’s pivot toward AI is reshaping operations—from clinical documentation to company-wide productivity—and why curiosity and hands-on experience are now essential traits for AI-ready teams.

Key Discussion Points & Insights

1. The Impact of AI on Clinical Trials

• Early Design Decisions Yield Massive Savings
  • Applying AI at the outset of clinical trial design can save upwards of $100 million in downstream costs.

    "By making judicious design decisions at the very beginning of the clinical trial, you can save upwards of $100 million in the downstream cost of the trial." — Patrick Leung [00:00]

• Automating Tedious Documentation
  • AI reduces the time to produce complex regulatory documents like clinical protocols from weeks or months to as little as 20 minutes for a first draft.

    "We're using AI to greatly accelerate that process from a matter of weeks or even months down to potentially, you know, 20 minutes to get the first draft." — Patrick [04:07]

• Combating 'Eroom's Law'
  • Drug development costs have risen inexorably, but AI promises to reverse this trend ("bending Eroom's Law downward").

2. Real-World Adoption of Generative AI in the Enterprise

Internal Productivity & Processes

• End-to-End AI Integration

  • Faro Health aggressively implements AI across the business:
    • Engineering: code prototyping, unit test creation, rapid API integrations.
    • QA/DevOps: automated testing, infrastructure-as-code.
    • Non-technical teams: sales teams use AI for research, designers for content and slides.

    "We’ve been really aggressively kind of adopting AI all over the place in our organization." — Patrick [10:02]

• AI Accessibility Has Changed the Game

  • The current wave of generative, multimodal AI is more transformative than the dot-com, crypto, or image-recognition booms because non-technical users can access its power.

    "It’s different from any other technology revolution I’ve been involved in." — Patrick [11:49]

The Voice AI Revolution

• Synthetic voice output is commercially available and lifelike, but Daigle notes current limitations in latency and responsiveness compared to advertised benchmarks [09:07].

3. Scaling Teams & Talent Strategy

• Challenges in Hiring True AI Expertise

  • Many candidates inflate their AI experience. Rigorous, practical coding tests are essential to screen for real capability.

    "It was just surprising ... the number of candidates that kind of fell down. And I didn't even really consider the coding test to be that hard, but it was kind of like you couldn't fake it." — Patrick [18:16]

• Lessons for Non-Technical Hiring

  • Use the equivalent of practical tests for sales, HR, or other functions to verify AI fluency, not just resumes.

    "Identifying a few tests, essentially the equivalent of a coding test, but for HR, for sales, whatever it is ... I think that's a fantastic idea." — Chris [19:10]

• Upleveling Through Curiosity and “Learning by Doing”

  • Rather than relying on certifications, Faro encourages tackling real problems with AI, fostering a “fail fast, learn fast” culture.

    "We basically said, look, we have these really challenging problems ... let's go do some research, let's just learn by doing." — Patrick [22:35]

4. AI Limitations & the Path to AGI

• Current LLMs Are Not on the Brink of “General Intelligence”
  • Patrick addresses hype vs. reality: LLMs can’t yet solve creative or high-complexity problems outside their training data. Real mastery comes after ~50+ hours of hands-on use.

    "I am not in the camp of people that think that large language models before too long will be AGI...." — Patrick [12:30] "Using AI to code is kind of a detriment for the most part until you hit about 50 hours." — Patrick [29:01]

• Divide & Conquer/Agentic Architectures
  • To overcome LLM limits, complex processes should be broken into smaller, manageable pieces coordinated by “agentic” systems—mirroring agile methods [32:45].

5. AI Governance & Security

• Enterprise Concerns: Data Privacy & Model Safety

  • Enterprise customers demand strict assurances that sensitive data won't leak into public models or cross customer boundaries.

    "Every customer we talk to, is asking what happens with my data? Is it going to be mixed with our competitors data? Is it going to be safe?" — Patrick [37:03]

  • Prompt injection attacks are an emerging threat, requiring robust governance models and constant vigilance.

• Advice for Evaluating Providers

  • Research privacy policies of vendors like AWS, Azure, OpenAI, Anthropic, and Google. Use LLMs themselves to interpret policy language if needed.

    "It's incumbent upon you to really research like look at the providers... look at what their policies are, look at what they commit to and make your own decisions." — Patrick [43:09] "Take those policies and review them in a large language model and ask here's my concerns based on this policy." — Chris [44:08]

6. Executive Insights & Industry Trends

• AI-Enabled Roles Beyond Tech

  • Chief Security Officers and Chief Product Officers are particularly impacted by AI trends, with responsibilities for governance, risk, and product vision [35:01].
  • Every executive needs to explore how AI impacts their area—those who don't risk being left behind.

    "At this point every executive should be thinking about how they can be using AI ... because if you won't, then your competitors will be." — Patrick [36:41]

• Evolving Industry Maturity

  • Attendees at conferences and industry events are asking more sophisticated questions about LLMs, evaluation, and avoiding hallucinations [26:27].

7. Recommended Practices for Teams & Leaders

• Foster Curiosity & Experimentation

  • Professional advancement—and organizational AI adoption—depends on a culture of curiosity and purposeful experimentation over static learning or credentials.

    "Curiosity is one of the key traits that determines how far you go in your career. So it's not just about large language models, it's about life." — Patrick [51:25]

• Don’t Blindly Trust Model Output

  • For code or strategy, always vet AI output against defined specs. Different models excel at different tasks—try several to find the best fit for each use case [47:01], [48:00].

Notable Quotes & Memorable Moments

• "We're using AI to greatly accelerate [clinical trial documentation] from a matter of weeks or even months down to potentially, you know, 20 minutes to get the first draft." — Patrick [04:07]
• "It’s different from any other technology revolution I’ve been involved in." — Patrick [11:49]
• "There’s only so far you can get by analyzing web pages and even sort of books out there. Like, I think we've already got to the point where a lot of the useful information has been processed." — Patrick [12:48]
• "Hiring is the most important thing you do really, when you're building a company. So it's worth the extra effort and investment and calling in some favors maybe even to do that." — Patrick [21:33]
• "Using AI to code is kind of a detriment for the most part until you hit about 50 hours." — Patrick [29:01]
• "Divide and conquer. And we've been doing that with very great results at Faro Health." — Patrick [33:47]
• "...if your system handles those crown jewels, then you have to be really sure that they're going to be safe." — Patrick [40:50]
• "Curiosity is one of the key traits that determines how far you go in your career. So it’s not just about large language models, it’s about life." — Patrick [51:25]

Timestamps for Important Segments

• Clinical Trials Transformation Through AI: [00:00] – [06:48]
• From Google Duplex to Faro Health: Patrick’s Journey: [02:18] – [03:44]
• Automating Documentation – Real World Impact: [04:07] – [06:48]
• Voice AI: What's Ready and What's Not: [07:52] – [09:07]
• AI Adoption Across Faro Health: [10:02] – [12:08]
• Why AGI is Still Far—And What LLMs Actually Do Well: [12:30] – [15:14]
• Reality Check on “AI Talent” & Upskilling Teams: [17:05] – [22:35]
• Parsing PDFs and Entity Resolution—Complex Enterprise Challenges: [24:09] – [25:47]
• Evolving Industry Maturity at Conferences: [25:55] – [27:23]
• Hands-On Learning is Essential for Teams: [29:01] – [32:09]
• Agentic System Architectures—Solving Complex Problems: [32:45] – [33:56]
• Who Owns AI Responsibility at Executive Level: [34:55] – [36:51]
• AI Governance and Upskilling for Security: [37:03] – [38:58]
• How to Assess Vendor Security and Data Practices: [39:36] – [44:08]
• Advice for AI Leadership and Cross-Team Guidance: [45:01] – [47:01]
• Choosing the Right Models For The Task: [47:01] – [50:37]
• Why Curiosity Matters—in AI and Life: [51:25] – [52:29]

Takeaways for Business Leaders

• Actively experiment with AI in core and non-core business functions; hypothesis-driven “learning by doing” outpaces traditional upskilling.
• Use practical, role-based tests to hire and assess for true AI fluency.
• Break down complex workflows into agentic, manageable chunks—AI is still best at sub-tasks, not entire business processes.
• Stay current on vendor privacy and governance, and leverage LLMs to interpret policy materials if needed.
• Foster a culture of curiosity—it’s the most future-proof trait for AI transformation.

For more from Patrick Leung, connect via the show notes or reach out to Faro Health. For AI readiness assessments, visit chiefaiofficer.com.