I'm Aniket Behera, co-founder of 100ms.ai, where we’re building voice AI for healthcare operations. In this episode of the 100ms Life podcast, I had the chance to sit down with Trellis Usher, founder and CEO of Nova Health Labs, a digital health startup focused on oncology.

Trellis brings a rare dual lens to this conversation: she’s both a breast cancer survivor and a longtime workforce transformation leader. Her perspective is uniquely empathetic, informed not only by years of guiding enterprise change at companies like Salesforce and Pax8, but by navigating the very healthcare system she’s now working to improve.

On our side at 100ms, we’re bringing our roots in real-time audio and video systems—from building infrastructure at Disney and Meta, to support millions of minutes of telehealth and voice AI interactions every week. This podcast was a conversation about solving real clinical bottlenecks with voice that feels fast, smart, and human.

From Patient to Innovator: Why Trellis Built Nova

Trellis’ journey into healthcare innovation started not as a builder, but as a patient.

“I was diagnosed with triple-negative breast cancer after a casual conversation with friends led me to finally schedule a mammogram,” she recalled. “When I got that call, my world shifted.”

As she moved through her treatment, Trellis encountered caring clinicians inside a system full of uncoordinated, fragmented operations. The toughest parts weren’t always clinical.

“The hardest part wasn’t the care itself, it was getting to the care. It was voicemails, waiting for callbacks, figuring out if something was urgent or not.”

That experience became the seed for Nova Health Labs, a company built to amplify empathy and reduce administrative strain for both patients and care teams.

Beyond the Chart: The Human Moments That Medicine Forgot

Trellis’ experience as a patient makes one thing very clear: clinical touchpoints are only a fraction of the patient journey.

“A patient’s experience is not limited to the care they receive in the hospital. It’s everything, the voicemails, the delays, the systems they have to navigate just to get care.”

Trellis pointed out that for many patients—especially those experiencing pain, fatigue, or physical side effects—apps and portals simply aren’t usable.

“If I’m going through chemo and my fingertips are numb from neuropathy, I’m not typing into an app. But I can still talk.”

Aniket expanded on that thought:

“Voice is the closest thing we have to human conversation—it carries urgency, empathy, emotion. That’s what makes it so powerful for care delivery.”

That belief drives how we build at 100ms, from latency optimization to audio resilience, to make voice-based experiences feel seamless and human.
Healthcare leaders know this intuitively, but don’t always act on it. Today’s patients engage with voicemail inboxes, callbacks, disconnected portals, and front-desk handoffs before they ever speak with a clinician. And for those with complex conditions like cancer, every delay increases anxiety, and sometimes, risk.

The Oncology Breaking Point: Why Patients Can't Wait for Solutions

Oncology teams are under mounting pressure. While earlier detection and longer survivorship are good signs, they’ve also outpaced growth in the clinical workforce.

“The supply of doctors, nurses, physicians, and really just capacity, is not keeping pace with patient volumes,” Trellis said.

One oncology nurse Nova spoke with shared how her day starts with 40+ voicemails, many from the previous afternoon. She said, “I feel like I’m already behind the eight ball before I’ve even started.”

And here’s what makes it even more challenging: patients often downplay their symptoms to avoid being a burden, unintentionally hiding the most urgent needs.

“Patients are saying, ‘I don't want to bother them,’ and that’s how we end up with urgent voicemails missed for hours or days,” Trellis explained.

This makes it hard for care teams to prioritize, and dangerously easy to miss red flags. For operations leaders, this is more than just a staffing issue, it’s a system design failure.

AI as Clinician Ally: Protecting the Time That Matters Most

One of the most important reframes in this episode is that AI isn’t coming for jobs, it’s coming for workflows.

“Nurses and doctors don’t go into medicine to sit on the phone all day. They go in to provide compassionate, clinical care, and we’re taking that time away from them with inefficient systems,” Trellis shared.

Her mother, a nurse, once told her she’d “fallen out of love with nursing” because it had become more about forms than people. That loss of purpose isn’t just a personal story, it’s a broader workforce risk.

Rather than removing the human touch, AI has the potential to bring it back, by letting clinicians spend more time doing what only they can: building trust, offering empathy, and making expert decisions.

“Let’s not be afraid of it. Let’s train it, refine it, and use it to do the work that clutters up our day—so we can keep doing the work only humans can do.”

Why Specialized AI Outperforms General Models in Healthcare

Many healthcare organizations are tempted by general-purpose AI, but Trellis emphasized why specialization matters. Nova is built exclusively for oncology.

“You don’t go to your PCP for a cardiac issue, you see a cardiologist. AI should work the same way.”

Each call helps the system learn, get smarter, and identify patterns. By training on oncology-specific language and workflows, Nova’s voice agent can triage based on risk, identify concerning patterns, and integrate seamlessly into EMRs, without overwhelming staff. That kind of specificity creates more value, more quickly. It’s a reminder for strategy leaders: depth matters more than breadth when outcomes are on the line.
At 100ms.ai, we take a similar approach. Rather than building generalist AI, we’re focused on delivering real-time, scalable voice infrastructure that partners like Nova can layer their clinical intelligence on top of.

As Aniket shared,

“We’re not trying to solve oncology workflows ourselves, we’re enabling the teams that know them best to solve them, with better tools.”.

That infrastructure-first mindset ensures performance, security, and adaptability across specialties.

Liberating the Healers: How AI Targets Paperwork, Not People

One of Trellis’ most powerful takeaways was this: AI works best when it evolves with, not apart from, the people using it. That’s why nurses, care coordinators, and front desk teams should be co-creators of the tools they use.

“They’re the ones closest to the inefficiencies. If we want AI that actually works, we have to build it with them, not just for them.”

This means involving staff not just in deployment, but in design. Because when the workflows improve, morale and care quality rise together.

At 100ms, this is something we see across specialties. The real value of AI is in automating what slows clinicians down, not what defines their job.

“Let’s not be afraid of it. Let’s train it, refine it, and use it to do the work that clutters up our day, so we can keep doing the work only humans can do.”

The Heart of Innovation: Building a Healthcare System That Actually Cares

Healthcare leaders are under pressure to reduce costs, improve access, and protect their clinical teams. In this conversation with Trellis Usher, six insights rose to the top:

1. Patient experience extends far beyond the clinic
2. Oncology is experiencing a dangerous capacity gap
3. AI must protect, not replace human expertise
4. Specialization is essential for real AI impact
5. Voice AI can deliver scalable empathy
6. AI must evolve with clinical teams, not around them

At the center of all this is a bigger question: Can we design systems that work better for the people inside them, patients and staff alike?

Our answer is yes. And companies like Nova Health and 100ms.ai are building for that future.

🎧 Listen to the full podcast here.

What's Next?

If your organization is exploring AI solutions, now is the time to start. Consider how 100ms.ai's AI voice agents can streamline your operations and unlock new efficiencies in your healthcare practice.

Follow 100ms.ai for continued insights into the intersection of AI and healthcare innovation. Or book time here.