Kamya Elawadhi, Co-founder and President of Doceree, is a healthcare marketing and business strategy executive who has played a central role in the company’s growth since joining in 2019. She has progressed through leadership positions spanning platform strategy, corporate development, client services and global partnerships, helping translate Doceree’s technology into measurable value for pharmaceutical brands, agencies, publishers and health systems. In her current role, Elawadhi leads the company’s global growth agenda and supports its expansion in the United States, with an emphasis on privacy-conscious and clinically relevant communication. Before Doceree, she held senior client and account leadership roles at HS Ad India and McCann Worldgroup, where she worked on healthcare, medical device and consumer brands while leading cross-functional and multicultural teams.
Doceree is an AI-first healthcare marketing operating system designed to help life sciences organizations engage healthcare professionals and patients with greater context and precision. Its platform combines verified professional identity data, real-time clinical intent signals, programmatic activation, point-of-care engagement and an AI reasoning layer that coordinates messaging across channels. By connecting physician awareness and prescribing decisions with the patient’s path to treatment, including affordability, initial prescription fills and refills, Doceree aims to give healthcare marketers a more unified view of campaign performance while supporting privacy and regulatory compliance.
You joined Doceree during its first year and played a central role in building its US presence before becoming Co-Founder and President. What initially drew you to the company’s founding vision, and how has that vision evolved as AI has become more important to healthcare?
What drew me to Doceree was the clarity of the problem being solved. Healthcare marketing had been running on borrowed infrastructure for years, tools built for other industries and applied to one of the most regulated environments in the world. The founding vision was to build something purpose-built for healthcare from the ground up.
What has evolved is the scale of what AI makes possible. Early on, the focus was precision: reaching the right physician at the right moment. AI has expanded what that moment can look like, from delivering a message to reading intent, initiating conversations, and connecting engagement across an entire care journey. The core belief has not changed. The ambition has grown because the tools have.
Doceree recently introduced Generative AI Connectors for ChatGPT and Claude. What capabilities do these integrations provide, and which healthcare or pharmaceutical teams are most likely to benefit from them?
The integration with platforms like ChatGPT and Claude is about bringing Doceree’s clinical and commercial intelligence into the tools that pharma teams are already using day to day. Rather than switching between systems, brand and agency teams can now query campaign performance, access HCP engagement insights, and surface actionable recommendations directly within their AI assistant workflow.
The teams that benefit most are those making fast, data-dependent decisions: brand managers tracking campaign performance across markets, agency teams monitoring HCP engagement patterns, commercial leads who need synthesis across multiple data streams without waiting on reports. It removes a layer of friction that has always existed between data and decision. That, for me, is where the real value sits.
How do the Generative AI Connectors fit into Doceree’s broader AI strategy, and what capabilities do you expect to add as enterprise adoption grows?
The connectors act as a layer of Daily Command, sitting alongside Semmelweis, our pharma-specific reasoning model, and the Analytics Workbench. The thinking behind them is straightforward: pharma teams are already working inside ChatGPT and Claude every day, so we bring Doceree’s clinical intelligence into the tools they already use, with HIPAA-safe guardrails built in from the start.
As adoption grows, the direction is toward deeper proactive intelligence. Systems that surface the right signal before someone thinks to ask for it. That is where we are building.
Doceree uses the term “clinical intelligence” to describe part of its technology. What does that mean from an architectural perspective, and how does it complement the capabilities of the underlying large language models?
Clinical intelligence, at its core, is the layer that makes general AI applicable to healthcare. Large language models are extraordinarily capable, but they do not natively understand a diagnosis code, a prescribing pattern, or what it means when a physician selects a specific therapy for a specific patient profile. Clinical intelligence is what bridges that gap.
Architecturally, it sits between the raw data signals coming from clinical workflows and the AI models interpreting them. It provides the context that makes outputs actionable rather than generic. Without it, you have a powerful model working with incomplete information. With it, you have something that can actually influence a real clinical moment.
How does Doceree’s Semmelweis AI system help interpret healthcare data and translate it into useful insights for pharmaceutical teams?
Semmelweis is built around a simple idea: healthcare data is only useful if it can be interpreted in clinical context, not just processed at scale. Pharmaceutical teams have enormous amounts of engagement and prescription data, but the gap has always been between having that data and knowing what to do with it.
Semmelweis closes that gap by reading signals across the clinical journey and identifying where intent is forming or fading. Those insights feed directly into Daily Command, our AI system of work for pharma brand teams, turning pattern recognition into recommendations that commercial teams can act on immediately, without needing a data science team in the room.
ChatGPT and Claude have different models, interfaces, and enterprise environments. What technical challenges did your team encounter when creating a consistent experience across both platforms?
The core challenge was ensuring that every prompt and response, regardless of which platform a user was in, stayed within the same compliance guardrails. Healthcare data does not flex based on which AI assistant is asking the question. The real work was building a governed integration layer that sits between Doceree’s clinical intelligence and both platforms, ensuring answers are always grounded in Semmelweis reasoning and Doceree data and always auditable end to end. Consistency in compliance was the harder problem to solve.
How do the connectors help users understand where an AI-generated insight came from and which data or reasoning contributed to the response?
Every response through the connectors is grounded in Doceree data and Semmelweis reasoning. When a brand team asks a commercial question, the response is backed by clinical intent signals, identity data, and campaign inputs specific to their brand. The entire workflow is governed and auditable end to end. In a regulated industry, that traceability is what makes AI actually usable in a commercial decision-making context, and it was a design principle from day one.
Healthcare organizations often work with sensitive data and complex regulatory requirements. How are privacy, security, access controls, and compliance incorporated into the design of these integrations?
This is something we thought about from the very beginning as a design principle. The connectors are built so that no patient data ever enters the workflow. Everything operates on de-identified HCP context only. Every prompt and response stays within HIPAA-safe guardrails, and the entire workflow is governed and auditable end to end.
What that means in practice is that a brand team can ask commercial questions inside ChatGPT or Claude and get answers grounded in clinical intelligence, without any of the data privacy risk that would normally come with connecting sensitive healthcare data to a general AI assistant. The guardrails are the foundation the integration is built on.
How do you evaluate the quality of the system’s outputs, including their accuracy, relevance, consistency, and usefulness for healthcare professionals?
Quality in this context means something very specific. An output is only useful if it leads to a better commercial decision, and that is ultimately how we evaluate it: whether acting on it moves a metric that matters, whether it lifts HCP stage progression, first fill rates, campaign performance.
The closed loop is what makes this possible. Daily Command is designed to feed outcomes back into the reasoning model, so every recommendation is sharpened by what actually worked. That is different from a general AI assistant where the output is evaluated in isolation. Here, the system learns from what happens after someone acts on a recommendation, which over time is the only honest measure of quality.
As pharmaceutical companies adopt generative AI, what principles should guide its responsible use in communications aimed at clinicians and patients?
The principle I come back to most is that AI in healthcare should narrow uncertainty, not create it. When a clinician receives information that influences a treatment decision, the source of that information and the reasoning behind it needs to be traceable. That is a basic standard of trust.
Beyond traceability, I think responsible use means keeping a human in the loop on anything that carries clinical or commercial weight. AI should recommend and surface. People should decide and approve. The moment that boundary gets blurred is when the risks start to outweigh the benefits.
And compliance has to be architecture, not policy. In healthcare, the difference between those two things is everything.
Thank you for the great interview, readers who wish to learn more should visit Doceree.
