Name: Pmsa Presentation 2026 3rr58ccolxw (1)
Uploaded: 2026-04-29T01:29:09.397Z
Duration: 19 min 37 s
Description: Pmsa Presentation 2026 3rr58ccolxw (1)

Transcript for "Pmsa Presentation 2026 3rr58ccolxw (1)": Good afternoon, everyone, and welcome to Comodo Health's webinar, deploying AI for quantifiable commercial results. I'm Carrie Dietz, a strategist here at Comodo, and it's my pleasure to introduce my colleague, Chad Forsey. With over fifteen years of experience leading commercial teams across health care and life sciences, Chad is a true pioneer in our field. In this era of rapid digital transformation, he is redefining how the industry views patient journeys and market dynamics through AI powered insights. Most recently, Chad has been instrumental in the evolution of Marmot, our groundbreaking AI platform. By fusing generative AI with the industry's most comprehensive health care map, he's empowering leaders to navigate market complexity and maximize their impact. Please join me in welcoming Chad Forsey. Thank you, Carrie, and welcome to all that are joining us virtually. We are opening this session with a clear mission, to move beyond the theoretical hype and to tackle the high stakes pressure that commercial teams are facing to implement AI that delivers actual quantifiable results. While adoption is accelerating across life sciences, the reality for most organizations is a frustrating landscape. A fragmented data and disconnected tools that are costly and nearly impossible to scale. This webinar explores what it truly takes to operationalize AI in a real world commercial environment, not in a boardroom or in a lab, but in the trenches where results are measured by patient outcomes and revenue. We are here to educate, not pitch, on the foundational role of AI ready data in enabling defensible decision making. Our goal is to move past the AI noise and find the agentic moments that matter. By the time we wrap, you'll have a practical framework to accelerate insights, reduce complexity, and drive consistent results. The analytics landscape is currently at a crossroads. While 70% of pharma leaders recognize the need to move faster, many are finding that generic off the shelf foundational LLMs are failing to deliver. These models are generalists. They lack the specific grounding and clinical nuance and medical ontologies required for high stakes analytics. When you are dealing with patient lives and commercial strategy, a general understanding isn't enough. Without the specialized grounding, these models don't just miss the mark, they can hallucinate insights that have no basis in clinical reality. This is exactly why your LLM experiments may not be giving you anything that you can act on. We are moving beyond the experimental phase to deliver trusted, evidentiary grade insights through a conversational data exploration that is native to healthcare. This is the difference between a generic experiment and an actionable business strategy. The do it yourself approach to AI frequently stalls because of significant data readiness gaps. DIY strategies force your most talented teams into a grueling slog of data engineering and integration. This can last for months and needs constant maintenance and configuration. All this just to reach a baseline for your analyses. You get trapped in a circular loop of aggregation, cleaning, normalization that drains resources before you've even asked a question. Our full stack advantage breaks this cycle by providing a vertically integrated platform that eliminates the need to stitch together fragmented data sources. This is what we call the Velocity Advantage. The ability to move from raw data to actionable insights in minutes, not months. By providing a pre built infrastructure, we allow you to skip the engineering slog and focus on strategic thinking, while directing agents to execute. This isn't just about speed, it's about reducing hidden costs and complexity that usually cripple an organization's attempt to scale AI. This advantage is powered by Komodo's healthcare map. We're capturing over 330,000,000 unique patient journeys. Unlike competitors who offer either data or tools, we merge together this comprehensive map with AI enabled technology onto one single platform. This provides the pre built infrastructure necessary to act as a force multiplier for your analytics team, supporting everything from clinical development to commercialization. This is a purpose built AI for healthcare analytics rooted in over ten years of pioneering data science and continuous refinement harnessing over a million unique analyses we offer stringently validated analytics that improve the trustworthiness of your data while actively reducing selection bias ensuring your faster decisions are also the right decisions. Now Marmot didn't start as a commercial product. It began as an internal engine designed to maximize skills across our own engineering analytics and commercial teams. As a health tech company, we've been using AI for years to support our products. And what we discovered as a company is that no single foundational model, be it GPT, Claude, or Gemini perfectly suited our needs. Each having its own unique quirks, its own capabilities. So the real breakthrough for Komodo was orchestration. Marmot orchestrates requests across this suite of models, selecting the optimal version for each specific task. As we are all in the exhaust of the industry's innovation, meaning as models version up, we incorporate those enhancements into our orchestration layer so you don't have to. The specialized infrastructure automates the heavy lifting of clinical translation, effectively compressing weeks of manual data engineering into a high velocity workflow that delivers defensible cone cohorts in minutes. By calling on specialized tools, like Research Planner or Cohort Builder, we've created an Agentic experience that finally allows you to talk to your data. This brings us to Marmot and our reference agent. Not merely a base model, but a highly optimized collaborative engine designed to drive your entire agentic workflow. Think of the Marmot Reference Agent as an advanced navigation system for the healthcare landscape. It is the sophisticated guidance core that knows exactly how to traverse the complex terrain of healthcare data with deep, fully optimized clinical context right out of the box. It provides the fastest, most reliable path from a raw question to a clear destination, Utilizing integrated tools that move with total technical precision, this is the mission critical core that collapses the distance between a raw clinical hypothesis and a defensible commercial strategy, empowering your team to optimize every agentic moment with true decision velocity. However, the extended value for your team lies in what happens beyond the core state. Marmot offers the unique ability to incorporate your institutional knowledge, your contextual data, your methods, your evaluations. You have the power to select specific tools from our scaffold, add your own datasets, custom prompts, or custom agents into the mix. This is where the experience becomes truly native to your organization. You can incorporate third party patient characterizations, layer in contextual data, and encode your specific business rules or user profiles directly into the Agentic framework. By writing specific context for its tasks, Marmot stops just being Komodo's AI solution and starts acting as a digital extension of your team's expertise. Crucially, this remains an enterprise grade experience that is not a black box. It was built for complete transparency, allowing any analyst to reason with the solution, interrogating the code, auditing the SQL behind every cohort and every drug list. It handles high volume heavy lifting so your experts can focus on strategic why questions, maintaining full confidence that the resulting analytic is research grade, auditable, and perfectly aligned with your internal workflows. Let's take a look at this colorectal cancer patient journey. Historically, analyzing treatment delays wasn't just a data exercise. It was a grueling multi month manual project. When we look at the bulleted list on the slide, we are seeing the anatomy of slog. In traditional workflow, the timeline for an analysis like this was dictated by heavy dependencies. In my past life, I might break this workflow into three phases. We'll start with the phase one, procurement and alignment, typically between one to four weeks. You start by defining research questions and desired insights, but you're immediately stalled by securing schedules, negotiating external contracts, and navigating the complexities of data access. Phase two, the engineering burden, typically weeks five through 12. This is where the emotional distress lives. We all know it. Your team must manually define HCP segments and map relevant ICD and CPT codes, often spending months cleaning and normalizing desperate data sources just to reach the baseline. And phase three, the iteration lag. And this is ongoing. Every time a new question arises, like wanting to stratify by age or metastatic status, the cycle restarts, and you're waiting for third party intervention or internal engineering windows to generate the next wave of insights. But with Marmon, we execute at the speed of thought because the agent understands clinical context natively. It can move from question to insight to strategic recommendation in minutes. The agent tracks events across time to pinpoint exactly where treatment bottlenecks occur, such as the 26% longer delay for older patients we see here. These patients wait, on average, thirty four days versus twenty seven days of a younger cohort. This isn't just a faster way to do an old task. It's a fundamental shift that puts 90% of a finished research grade analysis in your hands, allowing you to hypothesis test and pivot your strategy in real time without ever needing to leave the platform. Now this, this is what I call a gangbuster slide. And for an insights leader, this is where the ROI truly starts to compound. We've run one patient journey for Colorectal Cancer, but the intelligence derived from that single analysis doesn't live in a silo. It feeds the entire organizational engine. Instead of launching four separate manual consulting projects, you're now providing a unified source of truth that translates into direct impactful advantages for every segment of your business. For commercial, this we'll call Precision Revenue Acceleration. We move beyond the directional guessing to target specific HCPs, treating seniors 75 and older who are currently experiencing a 34 treatment delay. By identifying these exact bottlenecks, the impact is quantifiable. A simple seven day acceleration for just 15% of this cohort drives approximately 1,200,000 in annual net revenue. It's about focusing on the field efforts where the friction is the highest. And for HEOR and RWE colleagues, call it strategic research independence. Historically, quantifying unmet needs or racial disparities meant waiting months for external consultants. With Marmot, you can run these analyses internally, strengthening your payer and access discussions with research grade evidence. This isn't just a time saver, it's a cost driver, effectively eliminating 250,000 to $500,000 per year external consultancy spend while providing deeper, more frequent insights into underserved populations. And for Medical Affairs, a data driven care gap closure. The data reveals a stark reality. Seniors represent thirty three percent of colorectal cancer patients, but currently only eight point one percent of treatment starts. Medical Affairs can now prioritize high impact education at academic centers where the metastatic variation and treatment delays are most severe. You're moving from general awareness to targeted intervention based on where the care gaps are widest. And for clinical development, optimize trial intelligence. We pinpoint metastatic hotspots like the eighty percent rate we see in Utah, and you can optimize trial site selection in real time. This intelligence allows you to target underserved populations more effectively, reducing overall enrollment risk and building a stronger evidentiary foundation for future label expansion. By capturing the subjective nuance that matters to each of these stakeholders instantly, you move from four disconnected projects to one coordinated high velocity strategy. This is how a full stack platform transforms raw data into an enterprise wide strategic advantage. And now, let's talk about implementation. After moving beyond discovery, scoping, and contracting, we are hyper focused at getting this AI solution in the hands of the right users within your organization. Let's frame implementation as a three phase strategy. Phase one. We bring in highly technical and AI familiar users as first adopters. This allows our teams to understand ways we can transform your Marmot experience into one that natively maps to your organizational workflows. In phase two, we begin configuring your Marmon experience by incorporating your institutional knowledge, your contextual data, your methods, and your evaluations. And then in phase three, well, this is about adoption at scale. Finding the right users and the right teams that can act as a force multiplier for your brands and your business. And then the process starts all over again, continuing refinement of your experience and enabling more teams across your enterprise. Ultimately, we are solving for decision velocity. We had built a scalable AI infrastructure so your team can stop acting as data janitors and start acting as strategists. By moving to an AI ready full stack ecosystem, you unlock the I the ability to hypothesis test and to drive measurable results with a level of agility that was previously impossible. You are no you are no longer fragmented by data or disconnected tools. You are empowered to outpace conventional analytics and deliver consistent evidentiary grade results that drive commercial success and improve patient outcomes. And I'm gonna turn it back over to Carrie. Jeff, thank you. Thanks so much, Chad. It's clear the emotional distress of fragmented data isn't just a technical hurdle. It's a massive strategic bottleneck. Before we move to our final takeaways, I wanted to just address one last point that you made. The idea that we can finally stop being data janitors and start acting as strategists. I think this is the heart of what we mean by decision velocity. So before we close, I wanted to move into the q and a section. So, Chad, there's a few questions in the chat for you. I am gonna stop sharing my slides really quick. Few questions in the chat for you. The first one is, what is an agent? This is a great question. Right? I think in short, an AI agent is a system designed not just to answer questions, but to execute tasks autonomously. While a standard AI acts like an encyclopedia, just providing information, an agent uses reasoning to break down a goal and then select the necessary digital tools to then carry out a multi step workflow to achieve a specific result. We have agents on the back end that help through or a multi agent protocol essentially to answer these questions. We train them to understand the dynamics and context of our data as well as to code in a way that's really aligned to how health care analysts provide outcomes and results for our life science customers. Thank you. Okay. Next question we have is, what is AI ready data? Oh, another good one. Komodo's domain expertise with our data taxonomies, our ontologies, our semantic layers are the DNA of AI readiness. They transform an AI agent, like we spoke about before, from a blind chatbot that guesses into a digital colleague that understands. The health care map, a structured data product mapping The US health care system, a logical web of health care specific rules, a single source of truth applying health care specific definitions, this data framework allows a Marmon agent to reason with precision, to act with authority, and to navigate your business with zero hallucinations. Essentially, when an LLM provides the brainpower, an AI ready data product really optimizes the knowledge and the guardrails necessary for autonomous action that is safe and scalable and trustable. You. Okay. I see a theme. How do you ensure audit ready outputs? Yeah. That's a great question. When we set about building an agentic AI solution like Marmot, we acknowledge that for us, auditability is core. We would need to be able to trust but verify. And so, ultimately, what we've built is not a black box. The users can reason within the solution. They can confirm logic. They can see and edit the actual SQL code that's being generated. Not to mention, as a solution, Marmon is SOC two compliant. And we're also operating within the frame of health care analytics, so we take things like HIPAA and compliance very seriously. Awesome. And I think we have time for one more. How does this drive commercial performance? This really comes back to that decision velocity that you mentioned at the end, Carrie. I think getting insights faster to drive behavior, and then behavior then drives performance, and then performance really is unique to each team. Whether we're reducing spend or increasing patients treated, the the goal is for you to be able to talk to your data and go from raw question to a very compelling analytic output in a matter of minutes, and then to be able to hypothesis test and iterate. And then to take that information and to share it across the organization for that compounding ROI act effect so that none of this work, none of the lift that's done by your teams is living in a silo, and you're able to optimize the collective intelligence of your organization by leveraging an AI solution. So as we close today's session, we just wanted to leave you with a practical framework for identifying your own high velocity use cases. So first, look for this log. Where are your teams currently trapped in the data engineering cycle? Two, identify the ripple. Where could one high fidelity analysis like the CRC example we shared create a gangbuster ROI across your commercial, medical, and HUR segments? Third, audit your tools. Are you using generalist elements that hallucinate, or are you moving towards an evidentiary grade platform like Marmot that offers full transparency and auditable code? Thank you all for joining us today. Conversation, please feel free to reach out to Chad or I via LinkedIn, and we'll also include our contact information below. Thank you.