Table of Contents
You can read our first interview with Shashi Shankar, from March 2025, here.
Since we last spoke, could you give us an overview of how things have developed at Novellia in terms of patient growth, data scale, and enterprise adoption?
The biggest change since we last spoke is that we’ve reached a new level of trust, both with patients and with biopharma partners.
On the enterprise side, we’ve doubled the number of companies working with us entirely through word of mouth. We’ve done zero advertising, very limited marketing, and yet we’ve seen incredible pull from existing partners who are expanding and introducing us to others. Our customers span Top-20 pharma and several emerging biopharmas, which I think speaks to how valuable our solution is to virtually all types of biopharma companies. That growth has been organic because our partners see sustained patient engagement and data quality that improves over time, and not one-off datasets like you typically see with syndicated RWE vendors.
On the patient side, growth has accelerated alongside usage. In 2025, we grew our medical records exponentially with an increase of nearly 90x , and that number—now deep in the millions—continues to climb. Importantly, this is a story not purely about data volume. The real takeaway is Novellia users are now contributing richer, more longitudinal data as they actively use our product to manage their own health.
That combination—patients who stay engaged because the product helps them, and biopharma partners who benefit from higher-fidelity, continuously improving data—is what gives us confidence heading into 2026 as we scale more intentionally.
When it comes to deciding your internal direction, how do you balance your commitment to patient-centric data with the B2B partnership side?
First, we always, always prioritize what’s best for patients. It’s central to our mission and we believe it’s high time a data company put patients first. From there, it comes down to alignment. We talk a lot internally about the overlap between what creates value for patients and what creates value for our biopharma customers. That overlap in the Venn diagram is where Novellia exists.
Historically, many data vendors focused only on customer value and ignored the patient altogether. We flipped that. We ask: what do patients actually need? And we build from there. And what we’ve seen is that everything that’s good for patients delivers extraordinary value for biopharma.
Take symptom tracking. We did hundreds of user research sessions and learned that patients were struggling with managing symptoms. They would use their Notes app to log things like, “I had a headache at 3 p.m.”, or “sharp stomach pain at 4 p.m.”. So we launched a very intuitive symptom tracking tool without any marketing or ad campaigns. And almost overnight, it became one of the most-used features in the product.
For patients, that means clarity and continuity in their own care. For biopharma, it means access to structured, longitudinal insight into the lived experience of disease—outside the clinic walls. That’s the kind of real-world evidence that improves trial design, contextualizes outcomes, and ultimately supports better therapeutic decisions.
With the advent of AI, what proportion of your data comes from unstructured sources, and how are you using technology to digest it?
As our data sources have expanded, so has their complexity. Even within EMRs, schemas differ dramatically across health systems. Lab data introduces entirely different structures and taxonomies. Without significant normalization, that data is fragmented and difficult to interpret reliably.
Our AI pipelines are designed to solve that problem. We’ve built proprietary models that clean, harmonize, and organize disparate data into consistent, longitudinal patient records. For example, lab results that may be missing or buried in one system can be reconciled across sources to give a more complete clinical picture.
Because patients have consented to the use of their data, our models improve continuously as the dataset grows. That’s a critical distinction from traditional real-world data approaches that rely on de-identified or HIPAA-limited data and can’t be used to train proprietary models. Today, our systems are achieving accuracy levels in the low-to-mid 90% range, and improving with scale.
Does that give you an advantage in deciding which disease areas to focus on?
It does, because we don’t have to start with a narrow disease lens.
Our platform is built to be therapeutic-area agnostic. As long as a patient authorizes access, we can support a wide range of conditions—from broad populations to highly specific, ultra-rare diseases.
That flexibility matters for both patients and partners. For patients, it means their data can be organized and contextualized in ways that help them better understand their own health—what’s typical, what’s changing, and when to engage a provider. For biopharma, it means we can rapidly stand up disease-specific registries or population-level analyses without rebuilding infrastructure each time.
For example, today we’re actively generating actionable datasets on everything from breast cancer to ultra-rare conditions with only a few thousand patients globally. In many cases, we’re one of the only platforms capable of doing this at scale, because we combine consent, longitudinal depth, and analytical rigor in a single system.
Are there updates on future integrations like genomics or insurance claims?
Yes, and these integrations are driven by the same principle: giving patients (and researchers) a more complete picture of health over time.
We recently signed our first clinical genomics partnership in oncology. This includes variant-level data that can meaningfully inform therapeutic response and trial design. Historically, that information has been siloed or difficult to use alongside clinical data. Our normalization layer allows us to integrate it responsibly and at scale.
We’ve been piloting work on the claims side with a subset of patients. Claims data fills important gaps like confirming medication adherence, capturing procedures, and reflecting real-world utilization that EMRs alone can miss.
Together, these integrations move us closer to a unified, patient-authorized record that spans clinical, molecular, and behavioral dimensions of care. That continuity is what unlocks the next generation of real-world evidence.
How does your zero-fee patient model affect retention?
Our patient retention rate is over 99%, and that’s a direct result of trust and usefulness.
Patients don’t join Novellia to generate data: they join because the product helps them manage their health. Bringing fragmented records together, tracking symptoms, remembering medications—these are everyday problems, and we take them seriously.
By removing cost barriers, we make it easier for people to engage early and stay engaged over time. That sustained engagement benefits everyone. Patients gain clarity and control. Researchers gain more complete, longitudinal data. And the healthcare system benefits from evidence that reflects real lives, not isolated snapshots.
What does international expansion look like?
Many of our biopharma partners operate globally, and international expansion would allow us to support their efforts in a more integrated way. That said, we’re taking a very deliberate approach.
Patient consent and transparency are central to our model, which positions us well for regulatory frameworks like GDPR. Unlike platforms built on tokenized or de-identified data, we’ve designed our systems around explicit authorization from day one.
While it's early, we believe our strategy mirrors what worked for companies in this space before: establish trust, value, and operational rigor in the U.S., then expand internationally when it makes sense with a clear compliance foundation. When we do scale globally, we want to do it in a way that preserves patient trust everywhere we operate.
How are you scaling the backend and managing AI costs as you grow?
We’ve been disciplined from the beginning. Compute and infrastructure costs are tied directly to contracted work, not speculative usage. That ensures sustainability and strong margins as we grow.
We manage the full stack which allows us to optimize intelligently. Investments in areas like model compression and batch processing help control inference costs without sacrificing accuracy.
Efficiency is both a business and ethical consideration. We’re building infrastructure meant to last, and we believe responsible scaling is part of earning long-term trust.
How big is the team now, and how does it grow?
We’re a team of 16 today and expect to grow to 30–40 over the next year. Engineering remains our largest investment, particularly in AI, data infrastructure, and product. However, we're also investing aggressively in our commercial organization and quickly growing our sales org.
We’ve also expanded our executive team, bringing on leaders from places like IQVIA, Flatiron, and Symphony Health. As demand grows, we’re building biostatistics and analytics capacity to support more complex studies.
The goal is not growth for its own sake but rather depth, experience, and alignment around the mission.
Are you looking to move earlier into drug development, like phase 2 or discovery?
We’re already starting to support earlier-stage programs. That’s exciting because it gives the underlying dataset a much longer shelf life. It’s not just powering a single study, but informing evidence across the entire therapeutic lifecycle.
Discovery is a bit further out, but as we deepen our AI capabilities and expand our multimodal data, especially genomics and claims, we’ll be in a position to support applications like digital twins, virtual arms, or even early-stage target validation.
What about patient recruitment or direct-to-consumer pharma marketing?
Patient recruitment is a natural adjacency. It’s a crowded field, but we believe our ability to identify highly specific cohorts, often with longitudinal data and real-world behavior patterns, gives us a significant advantage. We’ve already had customers ask for help identifying likely responders or patients with specific diagnostic or treatment trajectories.
Direct-to-consumer pharma marketing is trickier. Most patients have a negative view of it (and understandably so!), and we’re not interested in spammy advertising. But there’s real value in thoughtful, educational outreach, like surfacing relevant trials or therapies that align with a patient’s unique profile. If we were to explore this, we would first thoughtfully invest in building strong guardrails around privacy and consent and reimagining this space in a more ethical, useful way.
How do you personally define success as CEO?
For me, success starts at the human level. It’s hearing from someone who used Novellia and it helped them. A family member of mine is a transplant survivor, and after trying the product, she texted me: “Holy s***, this is awesome. I don’t have to track my medical records or medications anymore.” And it’s not just the people I know. Complete strangers are using the product, and I’ll never meet most of them. That’s really motivating.
Beyond that, it's about impact at scale. If Novellia can accelerate how researchers understand therapies in the real world, or help a new treatment reach the right patients faster, that’s a massive success in my eyes.
And then there’s the team. Watching people come together around a mission, seeing them push boundaries and support one another, that’s been one of the most rewarding parts of this job. So much of startup life is hard. But when you're building something that genuinely matters, with people who care, that’s what makes the day-to-day worth it.
Looking five to ten years out, what does the patient ultimately get from Novellia?
Looking five to ten years out, the patient ultimately benefits from a healthcare ecosystem where their data finally works for them. Recent launches like ChatGPT Health demonstrate how powerful AI can be in helping people interpret health information and prepare for care - and that’s an important step forward. But AI is only as good as the data behind it.
At Novellia, our vision is to provide the foundational layer that makes those AI experiences truly meaningful over time: a complete, longitudinal, patient-owned health record that reflects decades of care, not just snapshots from individual visits. There’s simply no way for a doctor or an AI assistant to fully understand 50 years of health history in a 15-minute appointment without that continuity.
Soon, we’ll integrate with wearables, sleep trackers, and nutrition apps so patients can see how their lifestyle data and daily behaviors connect to their medical history, labs, and medications. Over time, that record becomes contextual, deeply personal, and incredibly powerful - not only for the patient, but also for clinicians and researchers working to cure diseases.
Doctors are constrained by limited time and fragmented information. Conversational AI tools help interpret what’s available in the moment. Novellia’s role is to privately and securely fill in the gaps by unifying health data across providers and over decades, with the patient fully in control.
We believe the future of AI in healthcare should be grounded in trusted, patient-owned longitudinal data that supports better decisions over a lifetime, not just a single doctor visit. And we believe that access to this kind of intelligent, longitudinal health record shouldn’t come with a price tag. Patients shouldn’t have to pay to understand their own health, and at Novellia, they don’t.
What questions should we ask Shashi next? Let us know in the comments.
Comments