How a Startup Practice Keeps 4 in 5 Patient Calls Off the Doctor’s Desk

New Hope Primary Care was built from the ground up as an AI and advanced-algorithm-first internal medicine practice. Before the doors opened, one question shaped everything: what should the front door look like? The front desk had to set patient expectations from the first call and stage the rest of the offering for success—so choosing the front desk became one of the most critical decisions in the entire patient journey.

On a recommendation from the team at athenahealth, New Hope chose Talkie.ai. Today roughly 4 in 5 calls are fully handled by the AI agent and booked straight into provider schedules. New patients enroll by phone more than through any other channel. And the physician-owner—who keeps staffing lean by design—fields only three or four calls a day, an order of magnitude fewer than earlier.

  • ~80% of calls fully handled by Talkie and scheduled directly to providers
  • 3–4 calls a day reach the physician—down an order of magnitude in weeks
  • #1 way patients schedule appointments—over portal and text
  • 24/7 coverage with no hold times, lunch breaks, or after-hours gaps

This embedded video is blocked by your cookie settings. to enable cookies and view the video.

New Hope Primary Care

Website:

www.newhopeprimarycare.com

1 location

in Columbia, MD

AI agent “Jessica”

“joined” the team in June 2026

Automated processes:

• Appointment scheduling
• Prescription refills
• New patient intake
• Clinical urgency triage
• Message for provider
• 24/7 after hours service

EHR:

Specialty: Primary Care

Designing the Front Door of an AI-First Practice

Most practices add an AI agent to relieve an overwhelmed front desk. New Hope had the opposite starting point: a new AI-first practice that had to decide what its front desk should be in the first place.

For founder Dr. Zeshan Rajput—an internal medicine physician and medical informaticist—the front door was a strategic decision from day one. It would set expectations with every potential patient and frame everything that followed.

When we came to the decision around what our front desk solution looks like, it was actually one of the most critical things we did in our patient journey, which is what brought us to Talkie.

Dr. Zeshan Rajput,
Owner, New Hope Primary Care

Read more about Talkie for Small Medical Practices

Why Talkie Cleared a Strict AI Policy

New Hope runs a strict AI governance policy: a human reviews every part of how AI is used, at every step.

We have a very strict policy around every portion of AI used here is human reviewed, human is in the loop at every single step of the way.

Two things made Talkie a fit. First, New Hope could work directly with Talkie’s development teams. Second, the oversight didn’t stop at go-live—Talkie provides a dashboard the two teams review together months later, watching real conversations the AI has handled.

Talkie also adapted to the practice’s regulatory reality. In New Hope’s state, Maryland, any recording must be disclosed upfront, and Talkie configured the experience to stay compliant with state and local rules alongside the practice’s own policy.

High-Reliability Workflows Inside athenaOne

The integration with athenaOne was a primary consideration—fittingly, it was an athenahealth employee who pointed Dr. Rajput to Talkie. Talkie files patient cases directly into athenaOne, slotting into the practice’s existing tooling and workflows.

When a patient asks for a callback within the hour, for instance, Talkie logs the case as priority one in athenaOne. That priority-one flag pages the on-call provider around the clock.

Talkie’s integration with athenaOne enabled us to create high-reliability workflows exactly where it matters for our practice.

Read more about athenahealth and Talkie integration

Patients Choose to Call; the Doctor Rarely Has To Answer

Despite having a portal and text messaging available, patients at New Hope keep reaching for the phone. Calling Talkie is now the number one way patients enroll and the number one way they schedule.

Even when we’ve made the portal available to them, we’ve made text messaging interfaces available to them, time and again patients are choosing on their own to call up Talkie.

Because the practice runs lean by design, the person picking up anything Talkie can’t handle is Dr. Rajput himself, or the on-call provider. So limiting the number of calls Talkie can’t resolve is one of the practice’s core success metrics. After tuning the system around the edge cases that were still slipping through, that number is now low.

I only field maybe three or four calls a day at this point, which is an order of magnitude lower than just a few weeks ago.

Always On, and Always Getting Better

A patient can call at three in the morning and get the same scheduling experience they’d get at nine. There are no lunch breaks and no after-hours gaps—and the system keeps improving with every conversation.

For an AI-first practice, that’s the front door working exactly as designed: setting expectations, handling the volume, and leaving the physician free to step in only where a human truly adds value.

There are no lunch breaks. There are no after-hours sessions. And these systems are constantly getting better.

Read more about Talkie for primary care practices

Ready to reduce your front desk staff’s workload?

Talkie’s AI agents will help you instantly answer 100% patient calls, and significantly free up your front-end team.

  • Reduce your staff workload and burnout

  • Automate up to 85% of the most repetitive calls like scheduling, Rx refills, and FAQs

  • Reduce call center operational costs by 60%

Book a Demo