Power Hour Optometry Podcast Summary

Episode Title: Building AI for Optometry: A Live Walkthrough of Practical AI You Can Use in Your Practice
Host: Eugene Shatsman (The Power Practice)
Guest: Obi Bharacharya (AI Consultant)
Date: October 31, 2025

1. Episode Overview

This episode offers a hands-on demonstration of building a practical AI tool to automate appointment reminder phone calls for an optometry practice. Host Eugene and AI consultant Obi guide listeners through the process, aiming to show just how accessible and impactful AI solutions can be for everyday optometric workflow challenges. The episode is designed for practitioners interested in leveraging AI but uncertain about where to begin.

2. Key Discussion Points & Insights

A. Why AI in Optometry Matters ([00:00]–[06:26])

• Eugene’s Perspective: AI is poised to improve patient experience, enhance practice efficiency, and generate new revenue streams.
• Common Barriers: Many practices believe adopting AI is hard, time-consuming, and costly. This episode seeks to challenge that belief by building a working solution in 30 minutes.

B. Obi’s Background in Health Tech and AI ([03:23]–[05:58])

• Obi explains the evolution from early algorithms (“if statements”) to today’s advanced generative AI.
• He shares experiences working with Athenahealth, Trusted Health (nursing workforce optimization), and Amazon Alexa—where he helped implement large language models to improve digital assistants.
• Quote: “AI is not new… What’s new is generative AI. Now AI can do pattern matching at scale.” —Obi ([05:36])
• He underscores the manual, repetitive nature of many optometry practice workflows—ripe for automation.

C. Live AI Build: Voice Appointment Confirmation Agent ([07:59]–[47:44])

• The Challenge: Automate patient appointment reminder calls using voice AI.

1. The Demo: AI Agent Calls Obi ([11:13])

• Obi triggers a call from an AI voice agent (built prior to the show) to his own phone, confirming his hypothetical appointment.
• Memorable Moment: Listeners hear the agent call Obi in real-time.
  • “Hi, may I speak with Obi?... This is Vision Care Optometry calling. I'm just reminding you about your eye exam appointment tomorrow at 2pm, can you confirm you'll be able to make it?” ([11:13])

2. Tool Stack Explanation ([12:28]–[16:19])

• Vapi AI: Handles the voice/phone interaction, allowing creation of voice agents that can make calls.
• N8N: A workflow automation tool that acts as the “brain” to manage appointment lists (from Google Sheets), trigger the calls, and log results.
• Google Sheets: Used to store and retrieve appointment data.
• Practical Tip: Don’t worry about full EHR integration at the start; practice automation with exported data instead.
  • Obi: “Don’t try to focus on the hardest parts of the problem first… Just focus on something small that you can get out that’s going to make an impact.” ([14:54])

3. Step-by-Step Build & Tool Chaining ([16:19]–[46:23])

• Using Claude (or ChatGPT): Obi shows how to use AI chatbots as “copilots” in the building process, starting prompts with a role (“Act as an expert AI/ML developer”) to get step-by-step guidance for setting up the automation.
• System Prompt Development: The system prompt serves as a “job description” for the AI—detailing tone, responsibilities, and guardrails, including patient confidentiality and what questions the agent can/can't answer.
  • "Think about this as hiring an employee... What would you tell an employee?" —Obi ([24:49])
• Voice Personality: Vapi AI allows choosing different voices, accents, and gender for your agent, making customization easy and engaging.
• Workflow Overview:
  • Pull appointment info from Google Sheets
  • Trigger the AI voice call via Vapi using the prompt and appointment data
  • AI agent makes the call to remind/confirm
  • Log the outcome back to the spreadsheet
• Iteration & Improvement: Calls are recorded/transcribed, allowing ongoing prompt refinement and troubleshooting using chatbots.

4. Live Test: AI Agent Calls Eugene ([47:52]–[49:23])

• Memorable Exchange: The AI agent calls Eugene live; he tries to reschedule, and the agent politely reaffirms the appointment and offers the contact number.
  • Eugene: “Oh, I thought it was at 2:30.”
  • AI Agent: “I have it scheduled for 2pm tomorrow. If you need to reschedule, please give us a call.” ([48:33])

5. Lessons and Practical Takeaways ([49:34]–[52:07])

• The AI agent can be incrementally improved, just like coaching a new staff member.
  • Obi: “This is like Eileen’s first day on the job… That would have been okay for a first call. I wouldn’t have fired her after that.” ([50:09])
• Evaluation and feedback are crucial—use transcripts, ask your chat assistant for improvement suggestions, and keep iterating.
  • “You can export all these transcripts, feed them to Claude and be like, Claude, improve the system prompt…” ([51:23])

D. The Value of Low-Code, Accessible AI for Practices ([52:07]–[55:35])

• The entire workflow is built using off-the-shelf, low-cost, and partly free tools, in less than an hour.
• AI can start with low-stakes, nonclinical workflows (such as reminders, billing, or marketing) before tackling complex clinical tasks.
• The collaborative build model: Use AI chat tools as your always-available “buddy” to help you brainstorm, troubleshoot, and implement solutions.

3. Notable Quotes & Memorable Moments

• Obi on Generative AI:
  “What’s new is generative AI… where AI now has the ability to do pattern matching at scale.” ([05:36])
• Obi on overcoming inertia:
  “Don’t let perfection be the enemy of done… Accept that this is going to be a testing and learning process.” ([09:15])
• Eugene on practical focus:
  “The most impactful part is, let’s figure out where we can save the time… How do we get this to do something that ultimately has an impact on the practice?” ([16:19])
• Obi on system prompts:
  “You want to give the AI a job description. It’s important to think about this as hiring an employee.” ([24:49])
• Eugene on iteration:
  “The cool thing about AI is that it’s not just software off the shelf. It’s like an employee you can continue to train and say ‘aha, in that particular case… let’s tell it what to do.’” ([49:34])

4. Timestamps for Key Segments

• Intro & Why AI for Optometry: [00:00]–[06:26]
• Obi’s Background & AI Evolution: [03:23]–[05:58]
• Setting the Episode’s Challenge: [06:26]–[09:15]
• Voicemail Agent Live Demo: [11:13] & [47:52]
• Tool Overview (Vapi, N8N, Google Sheets): [12:28]–[16:19]
• Step-by-Step Build: [16:19]–[46:23]
• Iterating, Evaluating, and Improving AI Agents: [49:34]–[52:07]
• Audience Call-to-Action: [54:09]–[55:35]

5. Flow & Tone

The episode is conversational, enthusiastic, and practical, with Eugene and Obi taking turns explaining and building solutions. The tone is encouraging and demystifying—emphasizing experimentation, iteration, and accessibility. Listeners are empowered to “play” with AI, understanding that even small workflow automations can yield practice-changing benefits.

6. Audience Engagement and Future Episodes

• Eugene asks for listener feedback and future build requests.
  “If you want to watch us build something... you could play with [these tools], or most important, you understand that AI can have practical application... Give us requests, and we’ll build something live on air next time.” ([53:04]–[54:09])
• The next episode promises to build AI requested by the audience—continuing the theme of real-time, practical demonstrations sourced from real-world optometry challenges.

Bottom Line:
This episode is a must-listen for optometry professionals curious about demystifying AI. It proves, step by step, that with widely available tools and guidance from AI chatbots, even small practices can implement time-saving, revenue-generating automations—starting today.