Marcello Damiani recalls lining up at a make-shift clinic to receive his first dose of the COVID-19 vaccine earlier this year. The clinic was set up at a Moderna Therapeutics plant in suburban Boston, which had been firing on all cylinders to make tens of millions of doses a month. Damiani was surrounded by fellow Moderna Therapeutics employees who helped design and test the vaccine, and would be getting a shot of their own product.
It was an emotional, anxious, and exciting moment for Damiani, who has been a central player in enabling Moderna to successfully design and produce products based on messenger RNA, a molecule that can hold “instructions” for training the body to fight viruses like COVID. (The company also has products in development for Zika, the flu, and cancer treatment.)
“You can imagine the mountains we had to climb, with this rapid scaling up, and the maturing of everything we had in place,” Damiani says. But in addition to feeling overwhelmed by the pressure and expectations, he and his colleagues were also excited, “because you could feel that the outcome of everything we were working on would impact you personally.”
Damiani is the biotech company’s Chief Digital and Operational Excellence Officer — a title that was crafted out of the company’s desire to apply information technology in a way that would make Moderna an agile and data-driven company.
“When I joined Moderna, we could have decided that I was called the CIO [Chief Information Officer] of the company. But what we focused on is how information technology is going to help Moderna improve its product, increase its scale, improve its efficiency, and improve drastically the quality [of products] that we’re building, and how we’re going to build a very data-centric company, and how we’re going to use algorithms to help the company do all those improvements.”
In a recent interview with InnoLead, Damiani shared insights about how that focus helped Moderna shepherd its COVID-19 vaccine to emergency approval in the US in December 2020; the challenges presented by the need to scale up the company’s manufacturing operations; his key principles when it comes to deploying new technologies; and more.
Can you talk about some of the things you had to do to support the creation of the COVID-19 vaccine?
Since 2013, we’ve been implementing digital solutions to support our scientists and how they proceed with ordering messenger RNA and designing their sequences and so on.
First, we used the cloud everywhere; everything sits in the cloud. We were, I believe, the first biotech doing that. We drive our manufacturing from the cloud. It provides huge flexibility should you implement a new facility, or a new line somewhere else. All you need onsite are the real-time, latency sensitive systems.
We developed what we call the Drug Design Studio… Our scientists…go to this [web] portal, and they select either existing messenger RNA sequences, or protein sequences, or they can design them from scratch. … They have access to all the chemistry that we have for them. At the end of this process…instead of going to the lab [to] pipette them manually, they push a button and send this data to the central automated preclinical manufacturing that we have in place. At every step of this process, and the preclinical manufacturing, you’re collecting data.
We have implemented lots of algorithmics to optimize the [mRNA] sequences for production. Some are easier to produce than others, so we use machine and algortithmics to suggest same sequence, but change a nucleotide, so that you can optimize your production.
We have implemented [all the data we have collected and algorithms we built] into the Drug Design Studio to help the scientists. We did the same on the automation side in preclinical. … We used many of those algorithms to build the COVID sequence, so we didn’t waste any time. If we had to do COVID in 2015, we would have failed maybe once or twice before we were able to produce it, because the sequence has its own complexity. But with everything that we built and the data that we had…we were able to produce it right the first time in the right quality…
What challenges did you run into while your production facility was being scaled up?
On the science side…I would say the scale-up was minimal, because we had everything in place. We were a research company, very focused, very digitized… We were set for [making smaller quantities of products for clinical trials], and we had to scale up Norwood, our facility [in Massachusetts], to [an] international facility with bigger volumes, more lines, and so on. But we had originally built it fully digitized and paperless. We had connected it with quality systems, so you can release the product only after you have quality reviews. Everything was in place. We used that setting to produce COVID vaccines – we had to scale up the quantities, the number of lines, but all the rest was already existing.
In 2015, we would have failed maybe once or twice…but with everything that we built…we were able to produce it right the first time.
But that’s only a small piece, because if you want to deliver on a global basis, you need to scale up all your supply chain. At the same time, we needed to build our capacity in Europe for international [distribution] using external partners…
We needed to work as well on our [clinical trial operations in 2020]… We worked a lot on diversity, making sure that we had a full representation of the population at risk in the US in our clinical trials. … Prior to COVID, we had integrated with the…systems [of] PPD, [a global contract research organization] who managed the clinical trials for us… Having this integration in place allowed us to collect real-time information about every site that was having a healthy volunteer coming, and real-time data about the population that we were recruiting… We corroborated this data with additional data from the public database about diversity — the address, population, and the area, and so on. This has helped us in real-time manage the clinical trials, adjust, and adapt to make sure that it was as representative as possible…
What was your strategy when it came to using data for projects like the Drug Design Studio?
The most important principle for us is integration. We wanted to make sure that all the data flows between the different systems… My experience from prior life was that many companies fail because they create silos of data. What they do is they put people and humans in between those silos, creating lots of inefficiencies translating data from one place to the other. That leads to very bad decision-making, because you don’t trust the data anymore. … We had a system of records that define, let’s say, HR data… We integrated the HR data into the Drug Design Studio, so we don’t have to keep all the names of the users inside Moderna — again — in the Design Studio…
The [next] part is around, can we integrate our instruments? Instruments generate lots of data. Usually what happens is that data goes into an Excel spreadsheet, and then you need to analyze them. … We integrated the instruments into the information system.
The holy grail is how you apply algorithmics and machine learning to this data. But we didn’t go for boiling the ocean, blue sky kind of projects.
The next thinking was, can we use robotic automation? … This is where operational excellence comes into play as well. … If you automate too early, and your processes are not stable, you slow down the changes because automation becomes very rigid. So what we adopted is, we start with islands of automation. We can connect those islands as the processes mature. As we interconnect them, afterwards, you have the full automation…
The holy grail is how you apply algorithmics and machine learning to this data. But we didn’t go for boiling the ocean, blue sky kind of projects. We took very pragmatic, specific problems we needed to solve, and looked for places where humans weren’t as good as machines.
Have any new best practices emerged over the past year that you now plan on continuing?
We had to mature our processes to make sure that they are consistent, reliable, robust before we can go out to the world. … As you can imagine, now that we have all those processes, whether it’s on the supply chain, whether it’s manufacturing, whether it’s on the commercial side now, they not only serve COVID. They will serve any other product that we are working on. That’s the beauty of the platform we have and how we built it.
We probably wouldn’t have not invested so much in digital and automation if we didn’t have this platform [for mRNA-based vaccines and drugs]. If you’re doing it for one product, the business case doesn’t stand up to invest as much in digital. The strategy was, we have a platform; we want to build multiple products; and let’s build it from Day One as digitized as possible.