It was mid-2025. Dr. Michael Dragutsky and I were sitting at his home in Memphis after dinner, playing around with AI. We ended up building a video avatar that could talk about colonoscopy screening. It took us a night to get started. Getting from that demo to an AI agent the front desk could work with like a remote employee took quite a bit longer.
“AI agents” had just entered the industry’s vocabulary. Everyone was still figuring out what to do with them. One GI was willing to go early. Their front desk teams gave us a wish list, and we picked fax referrals at Gastro Health Partners, a One GI practice.
The team had more patients waiting to be scheduled than staff had time to call. So, we at NextServices built an agent that could read faxes, call patients, and book appointments. We learnt a lot, even obvious insights such as patients would engage better during evenings and weekends when it’s harder to get humans to call.
As of this past week, that agent had booked 1,019 appointments. We estimate the value of the additional procedures scheduled by AI at over $500,000. I still find that number remarkable.
Along the way, NextDimension AI joined us as a partner, and we implemented an inbound AI call center. They surprised everyone by going live in two weeks flat.
This is the story of one of the first workflow AI rollouts in gastroenterology history. It may be the most detailed video case study we have in GI on what it takes to put AI to work in a practice.
We wanted to share it so other GI groups could learn from what we went through. What were staff nervous about? How would older patients react when an AI answered the phone? What would the physician partners say when patients started telling them about their calls?
Watch the 22-minute case study
In the video:
From One GI: Dr. Michael Dragutsky, Chairman, Carol Farris, and Kayla Kilburn
From NextDimension: Simran Khara, Co-Founder and CEO
From NextServices: Praveen Suthrum, Co-Founder and President
1. Chapter I: Delays in care make patients feel unimportant
2. Chapter II: “I was nervous about a full-time AI agent”
3. Chapter III: The secret to successful AI implementation
4. Chapter IV: The road ahead
Watch the video.

