How Ratio Therapeutics Built a Data Foundation That's Accelerating Scientific Discovery

Click here to listen now: Data Before AI: How Ratio Therapeutics Built a Data Foundation That's Accelerating Scientific Discovery on Edge of Excellence

Behind every successful AI initiative is something far less flashy and far more important: strong data fundamentals.

For many organizations, the challenge isn't adopting new tools but organizing their data in a way that makes those tools useful in the first place.

In a recent episode of Edge of Excellence, we sat down with Jacob Hesterman, Ph.D., Chief Data Officer at Ratio Therapeutics, to explore how his team tackled this problem from day one and why that decision is now paying off in ways they couldn't have predicted.

The Work Behind the Work

Ratio Therapeutics is a Boston-based oncology drug development company working in radiopharmaceutical therapy, a field where specially designed compounds are injected into the body, bind to tumor cells, and deliver targeted radiation to kill cancer. It's cutting-edge science that requires chemistry, biology, physics, imaging, and clinical medicine to work in lockstep.

Jacob is a physicist by training who came up through nuclear medicine imaging before pivoting into drug development. His CEO is an applied mathematician. His CSO is a radiochemistry and radiopharmacy expert. That multidisciplinary DNA shaped how the company approached its data strategy from the very beginning.

When Jacob joined Ratio nearly four years ago, the company had fewer than 10 people.

"Copy-paste can be the death of a lot of things," Jacob said. "It's amazing how quickly you get people working off individual spreadsheets in their own SharePoint areas, and then someone comes into a meeting and says, 'Well, this is what I have,' and somebody else says, 'Well, this is what I have.' That happens very quickly, even with a small team."

Rather than waiting for those problems to snowball, Jacob's team sat down with every scientific discipline in the company and asked a simple set of questions: What experiments do you run all the time? What are the standard inputs? The outputs? The metrics you care about?

That process became the foundation of Ratio's internal scientific data model.

Starting Simple and Building Smart

The early infrastructure started as structured tables in Excel, accessible, visible, and easy for everyone to understand. But it didn't stay there long. The team quickly introduced a relational database, then built what they now call the Ratio Analysis Platform: a web-based, read-only interface that gives anyone in the company (with appropriate permissions) access to that data in a safe, structured way.

One of their first real challenges was connecting data across disciplines. Ratio's biology team and chemistry team used different electronic laboratory notebooks. But both teams' work linked back to the same thing: a chemical compound at the top of a data hierarchy. By building API connections between those tools and structuring everything around that shared linchpin, the team created a single place where biologists and chemists could see each other's work without hunting through separate systems.

It was practical, unglamorous work, and it took time for the rest of the organization to see the value.

The Two-and-a-Half-Year Feedback Loop

Jacob is honest about the adoption timeline: it took roughly two and a half years before buy-in was truly universal across the organization.

"Everybody's busy. It's a startup. Everyone has a lot to do," he said. "And when some system kicks back and says, 'This is missing' or 'This is not quite right,' people are like, 'Do I really have to do that?' That's inevitable."

The other factor was simply the passage of time. In the early days, the team was small enough that everyone could hold most experiments in their heads. After a couple of years, that changed. Someone would ask, "Do you remember that experiment we ran?" and nobody really could. Now, Jacob's team can pull it up in 10 seconds, in exactly the same structure and format they always used.

One moment in particular crystallized the value. A scientist asked whether there was a relationship between mouse body weight and a specific output parameter across every study they'd ever run. Within an hour, one of Jacob's data team members had pulled up the complete experimental dataset and presented the answer. That was an eye-opener.

From there, a positive feedback loop kicked in. People started thinking: If you could do that, could you do this? The dynamic shifted from the data team chasing scientists down to scientists coming to the data team proactively saying, "We're adding a new assay. We want to make sure it gets into the data model the right way."

That subtle shift, as Jacob put it, made all the difference in the world.

Solving Pain Points People Didn't Know They Had

One of the more powerful stories from the conversation happened almost by accident. Jacob was walking past a colleague's office and saw them manually building a PowerPoint, pulling data from the analysis platform, pasting graphs slide by slide, and structuring tables by hand.

He asked what they were doing. Turns out, this was a weekly task for an entire team: assembling data into a standard format for a recurring meeting.

"I said, 'Well, that's something we can help with,'" Jacob recalled. "Now, Tuesday mornings, when people come in ahead of their weekly meeting, we kick off a job that automatically builds that PowerPoint in exactly that structure — for each of the programs the team wants to look at. They sit down at 8 a.m. with their coffee, and it's done. It saves them hours of manual effort every week."

That kind of automation is only possible because the data was already structured, governed, and accessible. Without the foundation, it's just a nice idea.

From Lab to Insight in Real Time

For a concrete look at how this all works in practice, Jacob walked through one of Ratio's most data-intensive workflows: biodistribution experiments. These are labor-intensive studies involving extensive animal work, measurement, and numerical processing.

Here's how the data flows: lab scientists take measurements on a scale, which automatically feeds into the electronic laboratory notebook. Once that data is complete, anyone with the right permissions can log into the Ratio Analysis Platform and kick off a job that runs all the required analysis steps. The system generates output spreadsheets and graphs, and because it's connected to Slack, those results automatically push to the appropriate channel, so the entire project team has real-time visibility.

Jacob's team even has observability into usage. They can tell when an exciting experiment is happening because multiple scientists will be hitting the tool simultaneously, refreshing over and over, waiting for results. "It's almost like a little view counter," he said. "We'll chat about it — 'Everybody's excited about something, because there's 12 different people trying to run this tool at the same time.'"

AI Readiness Is a Data Problem

When the conversation turned to AI, Jacob was characteristically measured; he's clear-eyed about what makes it actually useful.

The structured scientific data model that Ratio built over the past four years has positioned the company to take advantage of both traditional data science methods and newer AI capabilities. But they didn't stop at scientific data. Seeing the success of their scientific platform, the team applied the same principles to operational data — finance, corporate development, legal, HR — building API connections and creating a more integrated ecosystem across the entire organization.

At the same time, they recognized that not everything needs to be structured. For unstructured data like documents, presentations, and internal knowledge, they brought in a commercial tool that could index all of that content and make it accessible through a simple chat interface. That lowered the barrier to entry for people across the company who might not use the more technical analysis platform.

On the AI-assisted development side, Jacob's team has been pragmatic. For user interfaces and peripheral tooling, they lean heavily on AI coding tools. But for niche scientific and numerical calculations, they've found that more caution and human oversight is needed. "AI is often very confident in what it’s telling you," Jacob noted, "but a little more prone to error."

Advice for Organizations Getting Started

Jacob's advice for organizations looking to build strong data foundations comes down to a few key principles:

Start with communication. Sit down with every stakeholder group. Understand what they do every day, what matters to them, and what their pain points are. They're subject matter experts for a reason; your job is to understand their world well enough to build infrastructure that actually serves them.

Build progressively. You're not going from zero to 60 overnight. Identify the single most important rate-limiting step, tackle it first, and build sophistication one layer at a time.

Be honest about what's worth the investment. Not every data source or workflow needs to be brought into a structured ecosystem. Some things are better left as manual processes. Recognizing that distinction saves time and preserves credibility with your stakeholders.

Accept that you can build almost anything, but you can't build everything. When requests start rolling in (and they will, once people see the value), you need the ability to push back thoughtfully on what doesn't make sense right now.

The Bottom Line

The most powerful technology strategies don't start with tools; they start with foundations. Data governance, structure, and collaboration across teams create the conditions where innovation can thrive.

Ratio Therapeutics is proof that when you invest in fundamentals early, even when it's hard to see the payoff, you build the kind of infrastructure that doesn't just support your current work. It accelerates everything that comes next.

How iuvo Can Help

Building a data foundation that actually supports your team's work and sets the stage for AI takes more than good intentions. It takes the right strategy and the right partner. At iuvo, we help organizations design and implement the IT and data infrastructure that makes innovation possible. Let's talk about where you're headed. Visit iuvotech.com or reach out to start the conversation.

How We Create Our Content
As a future-ready technology company, we embrace AI as an accelerator to empower our teams and enhance the way we create. We believe that the reliability of AI technology depends on the people behind it, which is why every blog is supported by AI tools and then carefully reviewed, validated, and enriched by our subject matter experts. This balance enables and empowers our team to produce content that is useful, accurate, and trustworthy for our readers.