From Calls to Care: How Voice AI Helped Essen Healthcare Transform Care Coordination

At Team8, we believe the richest clinical data in the system isn't sitting in the EMR, it's flowing through millions of everyday conversations between patients and care teams.

In our earlier work on why voice + AI is coming to eat healthcare, we framed those conversations as an untapped, streaming dataset - the "gold rush" hiding in plain sight inside care management and population health.

At Essen Healthcare in New York City, that thesis is no longer theoretical. It's running in production, every day, on top of Popai Health -  a company we helped build together with Eyal Gurion, Elad Levy, and Michael Latar.

A safety‑net system under pressure

Essen Healthcare is one of the largest multi-specialty medical groups in the Bronx, with more than 600 providers serving some of New York City's most underserved communities. The organization started in 1999, when internist Dr. Sumir Sahgal opened a single primary care office in the South Bronx with a mission to serve underprivileged communities often overlooked by the healthcare system. What began as a one-internist practice built with an entrepreneurial spirit and dedicated mission has grown into an umbrella organization offering care for a vast array of medical specialty needs. 

Guided by their “We Care for That” philosophy, Essen's care coordinators sit on the front lines of value-based care, handling complex, high-risk patients, closing gaps, and managing a constant stream of inbound calls.

Like many health systems, Essen faced a familiar mix of pressures. Care coordinators were spending too much time on paperwork and documentation – time that could have been invested in patients instead. Quality assurance was almost entirely manual, which meant that across hundreds of staff members and thousands of calls, only a tiny fraction could realistically be reviewed. Leadership knew that every phone call held scheduling and follow‑up opportunities, but had little visibility into whether those opportunities were consistently captured.

Dr. Sumir Sahgal, Chief Medical Officer, Essen Health notes that, “after more than 25 years serving New York City’s most vulnerable populations, I’ve learned that many critical insights often come from the conversations we have with our patients. Popai’s conversation intelligence platform aligns perfectly with our ‘We Care for That’ philosophy by allowing our care coordinators to even further focus on what they do best: providing compassionate, high-quality care.” In other words: conversations were where the work happened, but almost none of that intelligence was visible in real time until partnering with Popai.

For Team8, Essen represents exactly the kind of environment Popai was designed for: high‑stakes, high‑volume, and under‑resourced – where the right infrastructure can change not just efficiency metrics, but the daily experience of patients and care teams alike. 

The question that shaped Popai

How do you turn phone‑based care coordination into a real‑time clinical and operational signal – without adding to the burden on already‑stretched teams?

Together with Eyal and his co‑founders, we anchored the company around value‑based care operations, care management, and population health – domains where thousands of calls a day drive risk stratification, outreach, and escalation decisions, yet much of the rich data and risk signals in those calls vanish the moment they end.

The core design principles the team set out to achieve were simple but demanding:

• Infrastructure, not a point solution — an underlying platform for "patient phone conversation intelligence," not a single workflow fix.
• Workflow‑embedded — documentation, insights, and triggers must surface in the systems clinicians already use, at the moment they can act.
• Clinician‑protective — every feature must give time back, reduce cognitive load, and respect the nuance of clinical reasoning.

Essen became one of the first places where this blueprint was tested at scale.

Bringing conversation intelligence to every call

Essen partnered with Popai Health to deploy conversation intelligence across 100% of patient phone calls in care coordination operations.

Popai's platform runs quietly in the background of every call, turning unstructured dialogue into structured, actionable workflows. Each call is summarized into compliant notes aligned with program and organizational templates, then written back into the systems clinicians already use. The same infrastructure identifies when an appointment should be scheduled, highlights latent scheduling opportunities that were missed in the moment, and brings them back into view for the team.

Risk is handled similarly. Calls that contain red‑flag signals — a worrying symptom, a medication barrier, a social determinant that could tip a patient into crisis — are automatically flagged for escalation. Those alerts don't depend on a coordinator's years of experience; they are standardized and consistent across the entire team. Supervisors gain a new lens on performance through real conversation data — seeing exactly where individual coordinators shine and where targeted coaching can shift outcomes.

All of this integrates with Essen's EHR and practice management systems, so insights flow directly into existing workflows rather than living in yet another dashboard.

What changed: time, quality, and control

Time back to care

Within months, Popai's implementation reached 92% adoption among more than 300 users — a remarkable rate in any large provider setting. Of those users, 88% reported measurable time savings, typically 25–100 minutes per day, returning 2–8 hours per week per staff member to the organization.

In practice, that means coordinators now spend more of their day on conversations, outreach, and follow‑through — and less time wrestling with notes and forms. For leadership, it's the equivalent of adding roughly two full‑time employees' worth of capacity without increasing headcount.

From a Team8 vantage point, this is exactly the kind of "top‑of‑license" shift we aim for in digital health: technology absorbing low‑value work so humans can spend more time on the nuanced, relational parts of care.

Capturing the details that used to disappear

With AI‑generated notes and structured extraction, 79% of users reported that Popai's documentation helped them catch clinical details they previously missed — including subtle but critical information like medication names and regimen nuances. Those captured details pay off in fewer clarifying follow‑up calls, stronger clinical documentation, smoother collaboration with downstream teams, and less friction around compliance. Notes now arrive in the chart consistent with program expectations, rather than reflecting each coordinator's personal style.

For us at Team8, this is a concrete example of an idea we return to across our digital health portfolio: when you respect the reality of clinical reasoning and patient narratives, and design infrastructure that can actually see that complexity, quality becomes a by‑product of better data — not just tighter checklists.

Running care coordination like a data‑driven operation

Because every call is analyzed, Essen now has measurable, transparent, and actionable care intelligence at its fingertips. Missed scheduling opportunities are surfaced and followed up on before the patient falls out of the funnel. High‑risk calls are cross‑matched with the EHR to detect gaps and trigger fast action — supporting better outcomes and reducing unnecessary readmissions. Supervisors have genuine oversight: they can see where guidelines are followed, where they aren't, and which calls merit coaching — without manually sampling a tiny fraction of interactions. In practice, this means next steps that would normally disappear inside a phone conversation are captured and acted on. As concluded in a case study written on how Essen transformed care coordination at scale, appointment scheduling became more effective as the system flags lost opportunities when they arise, resulting in more appointments proactively scheduled for patients.

This is the kind of "operating system upgrade" we look for when we back a company: not a new feature on the edge of the workflow, but a new way to run the core process.

The Team8 thesis in action at Essen

When we first partnered with Eyal and his co‑founders, our conviction was that voice would become a strategic data layer for healthcare. We saw three ingredients converging:

• The maturation of healthcare‑grade AI and infrastructure capabilities — a space where two of Popai's co‑founders, Michael and Elad, had already proven themselves at Aidoc.
• The shift to value‑based care and population health, where every missed signal carries financial and clinical weight.
• The sheer volume of phone‑based interactions that were going completely unmeasured.

Essen's results — high adoption, meaningful time savings, stronger documentation, more effective scheduling and escalations — are a live demonstration of that thesis.

They also reflect the broader Team8 model:

• Start with a well‑framed problem informed by deep market research and our Village of operators and clinical leaders.
• Pair it with a founder team that has true founder‑market fit — people who have lived the pain from the inside.
• Build infrastructure first, even if the first visible win looks like a "feature" (in this case, documentation automation).
• Prove the model with design partners who live at the sharp end of the problem — like Essen's care coordination teams.

Essen shows how that approach can move a health system from intuition to evidence on something as fundamental as the conversations their teams have every day.

Looking ahead: real‑time patient experience

Essen's next chapter focuses on patient experience in real time. Using the same conversation intelligence backbone, calls with outlier patient experiences are automatically flagged so a Patient Experience Specialist can intervene the same day — instead of months later when retrospective survey data arrives.

From our perspective at Team8, this is where the story loops back to the bigger picture:

• Conversations become a live sensor network for operations, quality, and experience.
• AI quietly translates what is said — and what is implied — into actions.
• Care teams gain the capacity and context to respond with more precision, not more burnout.

It's also a preview of where we believe digital health is headed more broadly: infrastructure that listens and learns continuously, surfacing the right signal to the right person at the right time.

From thesis to proof, and beyond

When we began working on Popai, we believed that "patient phone conversation intelligence" could become a new backbone for value‑based care. As concluded in the Popai - Essen case study, by implementing technology that seamlessly integrates with existing workflows, Essen freed care coordinators to focus on patient care while achieving measurable improvements in documentation quality, patient experience, and clinical outcomes. Essen's story is a concrete proof point: a large safety‑net system, 100% call coverage, high adoption, and measurable improvements in time, documentation quality, scheduling, and escalations. 

For us at Team8, it also validates something deeper about how we build: that with the right combination of thesis, infrastructure, and founder‑market fit, AI can be both ambitious and grounded — invisible to patients, trusted by clinicians, and relentlessly focused on better outcomes, better experience, and better use of scarce human time.

We're proud to be on this journey with Popai and with partners like Essen — and we're only at the beginning of what this kind of infrastructure can unlock for care teams everywhere.

Download the full case study here