What It Takes To Scale Care With AI | Akido Labs CEO Prashant Samant
Medicaid reimbursements are shrinking, providers are pulling back, and vulnerable populations are losing access to care. Akido Labs is betting that AI can expand care capacity fast enough to reverse that trend.
This week, Halle sits down with Prashant Samant, co-founder and CEO of Akido Labs, to discuss what it actually takes to scale care with AI. They explore why Akido built a full-stack healthcare company, how its AI operates inside real clinical workflows, and why the hardest patients are the best place to test whether this model works.
We cover:
Why he chose to build a full-stack care model
How AI changes who can deliver care, and where
Why most healthcare AI tools fail once they hit real clinical workflows
Why the doctor shortage cannot be solved by training more doctors
How the bottleneck in healthcare AI is absorption, not innovation
About our guest:
Prashant S. Samant is CEO and co-founder of Akido, a healthcare technology company that builds clinical AI and operates a multi-state medical network serving hundreds of thousands of patients. He co-founded Akido in 2015 through USC’s Digital Health Lab. In 2023, he and his co-founders received the EY Entrepreneur of the Year–Greater Los Angeles Award. Samant is also a co-founder and board member of Grid110, a nonprofit accelerator supporting early-stage entrepreneurs. He holds a bachelor’s degree in economics from Washington University in St. Louis.
—
Show Notes
—
Tune in
Full Transcript (AI edited for clarity)
Halle Tecco:
Hello listeners, and welcome back to the show. Today I'm speaking with Prashant Samant, the co-founder and CEO of Akido Labs. Prashant, welcome to the show.
Prashant Samant:
Thanks for having me, Halle. Great to be here.
Halle Tecco:
Your company runs a full-stack, AI-enabled medical network of 96 clinics that currently supports 500,000 patients. I want to talk about the AI-enabled part. What are you doing that's unique here?
Prashant Samant:
There are two parts to it. One is that we build our own technology from the ground up. The second is that we’re not selling AI into someone else’s workflow—we operate the care delivery system ourselves. That means we design the visit, the staffing model, and the follow-up process around AI from the ground up. Most healthcare AI makes existing systems a bit more efficient, and we’re using AI to expand care capacity itself.
Halle Tecco:
What is Scope AI? Can you walk us through how it works in a patient visit?
Prashant Samant:
Scope AI is a suite of neural networks trained on a large amount of proprietary data, along with a reinforcement learning system we’ve built in-house. It uses our electronic health record and takes information from visits that are quantified, qualified, and fed back into the system. You can think about it as a training environment where we’ve created one of the world’s largest medical textbooks, as well as one of the largest multi-specialty residency programs, to train a specific medical intelligence. Scope spans 26 different specialties and performs at an attending level of investigation and diagnosis in those areas.
Halle Tecco:
So how does that actually work in a patient encounter?
Prashant Samant:
In a Scope-enabled visit, a medical assistant, care manager, or community health worker is with the patient, and Scope helps drive the intake and clinical investigation in real time. It guides the conversation, identifies follow-up questions, pulls together the relevant history, generates the documentation, and proposes an assessment and plan. Then a physician reviews, edits if necessary, and approves. That’s the full cycle.
Halle Tecco:
Do you use your own EHR?
Prashant Samant:
Yes, we have our own EHR.
Halle Tecco:
So everything is trained on the data you’ve collected over time?
Prashant Samant:
It goes beyond that. In the first five years of the company, we focused on building a massive amount of organized clinical information. That didn’t really exist at the time, and in many ways still doesn’t. We integrated data across healthcare systems and social service systems, then stitched that together into a collection of over 10 million longitudinal clinical case studies.
Halle Tecco:
How did you get access to that data?
Prashant Samant:
We started at the USC Digital Health Lab, where we began working with health systems and public programs. We realized we needed not just clinical data, but social context as well. Over time, we built a synthetic dataset—a representative, anonymized population based on that underlying data. It’s like creating a textbook from real-world case studies, without using identifiable patient data.
Halle Tecco:
How did you know you could trust it?
Prashant Samant:
We validated it through real-world deployments. As we integrated data and deployed solutions in the field, we could test whether the synthetic population was representative enough. At scale, with around 10 million case studies, we could build something that reflects how people actually behave and respond to care.
Halle Tecco:
You don’t plan to sell the software?
Prashant Samant:
Not right now. In healthcare, the workflow is the product. If we just deploy software into existing systems, we won’t see the same impact as redesigning care around it. From a mission standpoint, this approach helps us drive change faster and more effectively.
Halle Tecco:
What have you learned from deploying this in real care settings?
Prashant Samant:
A big lesson is that technical capability and operational adoption are different problems. Sometimes the model is ready before the workflow is. You have to iterate in a live care setting and tighten that connection. That’s how you build something that actually works.
Halle Tecco:
How do your care teams respond to that kind of constant change?
Prashant Samant:
We focus on solving real problems they face. If the technology makes their work easier or more effective, it gets adopted. It’s not about forcing tools on them—it’s about building something that helps them in real time.
Halle Tecco:
Let’s talk about your street medicine work. Why focus on the unhoused population?
Prashant Samant:
If our model only works on easy patients, it’s not a healthcare breakthrough—it’s a convenience product. We wanted to focus on populations where the system fails most visibly. If it works there, you’re solving a real access problem.
Halle Tecco:
How does your care model work in that setting?
Prashant Samant:
We use a pod model with providers and community health workers. The goal is trust, follow-up, and coordination. Community health workers go into the field—tent communities, shelters—and engage patients directly, addressing both medical and social needs.
Halle Tecco:
How do you maintain engagement with a highly mobile population?
Prashant Samant:
We’ve seen over 80% retention, largely due to frequent touchpoints. With AI, we can maximize each interaction and extend the reach of our care teams.
Halle Tecco:
Are payers supporting this model?
Prashant Samant:
In some cases, yes, but reimbursement often lags innovation. The traditional response to Medicaid pressure is to cut capacity. We think that’s backwards. The only durable solution is to lower costs while increasing access.
Halle Tecco:
Do you think AI is the only way to prevent system collapse?
Prashant Samant:
I do. We can’t train our way out of the provider shortage. AI allows us to scale medical intelligence beyond individual human capacity while maintaining human oversight.
Halle Tecco:
Some worry this creates a two-tiered system.
Prashant Samant:
We already have a two-tiered system—people who get care and people who don’t. Increasing access is the most important step in reducing that gap.
Halle Tecco:
If you could change one thing about healthcare, what would it be?
Prashant Samant:
I would make the system able to adopt better care models as fast as technology makes them possible. The limiting factor isn’t whether we can build better tools—it’s whether the system can absorb and operationalize them. When it can’t, the people who suffer most are those with the least access.
Halle Tecco:
Prashant, thank you so much for your time.
Prashant Samant:
Thank you for having me.