My Career Building Platforms for Science

This is the story of my highly unusual and sometimes hard to believe career. I've told it to friends, family, partners, and people in the industry over the years as it was happening. Most often while reminiscing over a glass of wine. Eventually someone I respect a lot from my career cornered me and told me to put pen to paper and publish it (You know who you are). So I built this site to tell it. I hope you enjoy it as much as I have enjoyed living it. My goal is for you to learn something about my career journey you can’t find on LinkedIn.

Part I: Advanced Peptides (2007–2014)

The Beginning: From Lab Bench to Business Plan

I was fifteen when I started working as a paid scientist at the Whitehead Institute in Cambridge. My mother worked in the Whitehead libraries, which gave me the chance to meet people there. I had been attending their annual high school lecture series since I was twelve, and by fifteen I was taking science and math classes at Harvard alongside my high school APs. Alan Jasanoff, a young PI building his neurophysics lab, gave me a shot.

By sixteen, I was engineering and producing novel proteins, conjugating them to newly discovered nanoparticles in an effort to build next-generation contrast agents for fMRI. Not a typical high school science project. They threw me in, handed me textbooks and articles, and told me to figure it out if I wanted to stick around. That experience set my definition of what research science should be. It trained me to expect the highest bar, and it taught me to trust my abilities. In science I never had imposter syndrome. I had plenty of it in other parts of life, but never when it came to the work.

The Realization: Why Wait Until Forty?

By the time I was a junior at Boston University studying biomedical engineering, my plan was to follow the traditional academic track. Professor, running my own lab, doing research I could lead. One night with friends, we talked about what that actually meant. Not just a bench or a project, but a lab with autonomy, resources, and a team. We started counting the years like a checklist. Finish undergrad. PhD, five to six years. Postdoc, three to five years. Assistant professor. Associate professor. Then full professor, and only with tenure would I finally have control and resources.

I was twenty years old. By the time we finished, I realized it could take until forty before I had the independence I wanted. I happen to be forty now in 2025 as I write this. It felt like living an entire life over again before I could start doing the work I wanted. I was impatient.

I asked my father, who was on the research faculty at Harvard Medical School, if this was really the only path. He laughed and said, "That's if you're lucky." He told me I should do myself a favor and become a lawyer or an engineer instead, since that would be a better quality of life than a struggling academic scientist. When I pressed him for an alternative, he said, almost as an afterthought, "You could start a company." It was not encouragement, just a statement of fact. But it stuck.

Finding the Gap: The Birth of Advanced Peptides

At that point I had been working in protein engineering, conjugations, and chemistry for years. I kept noticing that many of the CROs our labs used were unreliable. We would send projects out, and they would either come back late, low quality, or not working at all. I would then take the same problem, do deeper research, experiment with different methods, and find ways to make it work. That was the gap. I realized I could start a company focused on the hardest peptide problems, deliver better science, and have universities and pharma companies pay for it. They would fund my labs, my people, and the profit, and I could reinvest that into programs of my own. That was the initial thesis for Advanced Peptides.

I spent the next year learning what I didn't know. Talking to business school professors. Learning to read a P&L. Meeting banks and investors. Trying to understand what it would take to run a science business. By September 2007, at twenty-two years old, I had secured funding to build a peptide lab. Looking back, I think I was lucky. It was right before the financial crisis. I still cannot believe investors trusted me with their capital at that age, with no real track record outside of technical skill, conviction, and confidence.

Surviving the Crisis: Trial by Fire

After months of searching, we found a place to start. An old ink and resin shop on Commonwealth Avenue in Boston. We gutted it and converted it into a peptide lab. Bought equipment, hired a small team, set up operations, and began looking for partnerships.

The reality of running a business hit quickly. Before the first year ended, the 2008 financial crisis hit. Research budgets were slashed. At the same time, a global acetonitrile shortage directly threatened peptide chemistry. I remember thinking to myself, "You wanted this," when everyone looked at me to find a way forward. We scrambled to redesign processes, find alternative vendors, and keep the company alive. That was the first of many ups and downs. Vendors failed. Supply chains broke. Cash got tight. You just kept figuring it out.

Scale and Innovation: 3,000 Projects Later

Over the next eight years, we worked on more than 3,000 projects with over 120 organizations. Universities like Stanford, Harvard, MIT, Yale, Princeton, Oxford, and A*STAR. Pharma companies like Pfizer, Merck, Johnson & Johnson, EMD, and Bayer. Our work grew from synthesis and conjugations into designing experiments with partners, producing molecules, analyzing results, and then into formulation, delivery, scale-up, manufacturing, GMP and GLP programs, preclinical and IND-enabling studies, clinical support, patent strategy, and in some cases product launch. Some projects were fee-for-service, some were co-investments, and some were built from proprietary scaffolds and ligands unique to us.

The Data Insight: Early Machine Learning in Drug Discovery

Between 2011 and 2013, after running thousands of projects, I began to notice a pattern in how our partners were using data. They would design 1,000 peptides, screen them through assays, move 20 forward, then 5, and discard the rest. The data was used only to decide what to advance. Almost no one was looking backward. What kinds of modifications correlated with binding? Which conjugations produced the folding or epitope activity they wanted? How could you mine prior results to improve the next round of design? That was the seed of a new idea.

We built a software platform that combined our project data with partner data, and used it to predict better peptide designs. It was one of the earliest applications of machine learning in this space. It turned out to be valuable enough that we spun it off and sold it to a close partner. That proved we could innovate beyond services and create technology that stood on its own.

The Exit: Time for Something New

By twenty-nine I had lived through more projects than most people see in twenty years. That foundation made everything I did next possible. And the truth is, I was fortunate. Most people didn't realize how young I was. Pre-Zoom, most interactions were by email and conference calls. My concepts showed up before my age did, and that gave me credibility long before I looked like I should have had it.

Around the same time, one of our joint projects with a multinational went commercial. A patent we co-developed was licensed, and for the first time I had stable personal income from royalties. That gave me the space to step back and decide if I wanted to continue down this path or build something new.

That experience still shapes how I think about drug development and how the industry works today.

Part II: SmartLabs (2014–2024)

A New Vision: The TSMC of Pharma

By 2014 I was looking for my next challenge. I met two people who became my co-founders: PC Zhu and Seth Taylor. PC had a PhD in Biochemistry from Fudan and an MBA from Chicago Booth. He ran a biologics and antibody CRO in a space adjacent to peptides. Seth had a PhD from Johns Hopkins and an MBA from MIT Sloan. He was running a consulting firm advising Fortune 200 pharma on R&D strategy and M&A. Compared to them, I was a decade or more younger, but I had a well-rounded mix of science, operations, and strategy from building Advanced Peptides.

Together we saw the same shift. Pharma was externalizing R&D into startups and mid-size companies at an unprecedented rate. At the same time, new therapeutic modalities were emerging: gene editing, mRNA, cell and gene therapies, viral vectors, protein degradation. Each required infrastructure, resources, and workflows that existing labs were not built to support. The outsourcing trend and the rise of new modalities were happening together, and they were going to reshape the industry.

We wanted to be part of that change. We believed the industry needed a new kind of platform. Not an incubator, but enterprise-quality labs with operations, compliance, and resources built in. We saw an opening to bring fractional infrastructure to pharma, the way it had transformed other high-tech industries. TSMC had done it for chip fabs. AWS had done it for data centers. Fractional infrastructure reduces cost, time, and friction. It lets industries move faster and innovate at scale. Pharma had never embraced it. The industry was expensive, risk-averse, and tied to dedicated infrastructure. With science moving into unfamiliar territory, we believed the time had come.

The Launch: From Vertex to CRISPR

We raised $8.8 million in seed funding in January 2015. Ian Smith, the CFO of Vertex, took a flyer on us because he believed in both our team and the problem we were trying to solve. He backed us with a 124,000 square foot sublease of Vertex's former R&D headquarters at 675 West Kendall Street. That building became our launchpad. We called the company Mass Innovation Labs.

Part of what reassured both Vertex and our investors was that we secured an operational partnership with Charles River Laboratories. Together we launched the Charles River Accelerator and Development Labs, CRADL. CRL filled in at launch. They stood up our first vivarium program, provided compliance infrastructure, and gave us credibility when it was still just three founders and a whiteboard. We in turn helped them finally build the CRADL vivarium hotelling model, an idea that had been debated inside CRL for more than seven years but never executed. They launched it with us on our site, while also supporting us with joint marketing. That partnership bridged the gap from an idea to a functioning business.

CRISPR Therapeutics was our first tenant. They moved in while we were still assembling furniture. In their first year, they grew from three people to fifty. We reconfigured space, sourced equipment overnight, and rebuilt offices over weekends to keep up with them. The early work they did in that building became the foundation for what later became the first FDA-approved CRISPR therapy for sickle cell disease. From our labs, they generated proof points, secured partnerships, and moved into the clinic at unprecedented speed.

The Problem: Static Labs in a Dynamic World

Within a year, we ran into our first structural problem. Demand was strong, but our inventory rarely matched what clients needed. If we had open space with fume hoods and a client wanted segregated tissue culture suites, we could not deliver. Even when the need was close, like a request for 15 fume hoods when we had 5, the mismatch killed the deal. Our labs were too static, and the build times and costs made no sense for a company only planning to stay three years.

I asked myself the basic questions. What is a lab? How does it actually work? Why is it built the way it is? And if it needed to change, what would you do differently? Every expert I spoke with said it was impossible. By definition, labs were fixed infrastructure. I didn't believe that. I had spent years on robotics teams at MIT in high school, building underwater vehicles modularly, swapping out components to solve new problems. Friends at Apple, Tesla, Google, and SpaceX were doing the same thing in their fields. If modular systems worked there, why not for labs?

The Breakthrough: Modular Labs

I convinced our investors to let me build a prototype. My pitch was simple: if we did not fix this core problem, our company would eventually be worthless. We tested the idea by ripping out a heavy chemistry lab with sixteen eight-foot fume hoods, space that had been dead inventory for months. We rebuilt it with a prototype modular framework and marketed it as "anything space." Within a month, we leased it to four very different companies: Beam Therapeutics, Oncorus, Edigene, and a battery technology startup. That was the breakthrough.

In November 2017, after proving the concept, we raised a $14.8 million convertible note to fund the transition. We began building our own system of modular walls, segregated utilities, HVAC, and digital controls. We also began developing software to manage logistics, bookings, and compliance. That eventually became SmartLabs OS.

Scaling Up: From Boston to National

In 2018, we rebranded from Mass Innovation Labs to SmartLabs. We partnered with BioMed Realty, Blackstone's life sciences arm, and opened a new 21 Erie Street facility in Cambridge, our first full building designed with the modular framework. It hit full capacity within six months. Around this time, CRADL and SmartLabs also went in different directions. Both models grew, but we had different ideas on how to scale the vivarium business.

In 2019 we raised our Series A-1, $36 million, which carried us through early expansion across Boston.

By 2020 we needed to move from three facilities to five. That was the purpose of our Series A-2, $45 million. With that we opened our first site in South San Francisco, an 80,000 square foot facility, and proved the model worked outside Boston.

In 2021 we raised our Series B, $250 million. That round was about scale. At that point we had roughly 300,000 square feet under management. Our goal was to triple it. The Series B was designed to take us close to a million square feet, a national platform instead of a regional one.

Enterprise Validation: Big Pharma Joins

In those years we also saw our client base shift. We had started by proving the model with startups and growth companies across Boston. By the time I left, about half of our clients were global pharma. The same enterprise groups that had always been the most risk averse were now relying on our platform. They adapted, and they benefited. That validation showed that this model wasn't just for emerging companies. It worked for the largest and most established organizations in the industry.

By the end of 2022 we had about 450,000 square feet open and operating, with another 500,000 square feet of future growth planned and committed to. We had also launched pilot-scale cGMP facilities on both coasts, extending our model into manufacturing. The bets we put in motion in 2022 were designed to carry us through 2023, 2024, and 2025.

The Market Collapse: 2023

Then the market collapsed. The downturn in 2023 was sharper than anything I had seen in my career. IPOs disappeared almost overnight. Private capital for pre-revenue companies dried up almost completely. Almost half our clients were affected, many of them the very startups that had been driving the new modalities forward. The growth bets we had made, new facilities, new capabilities, new geographies, suddenly ran headlong into a market that was no longer there to support them. Instead of planning for scale, we were watching once well-funded clients cut programs, shrink teams, and in some cases shut their doors entirely. It was the most dramatic shift in sentiment and funding I had ever witnessed, and it hit directly at the foundation of our business model.

We pivoted. We stopped new expansions, pulled back from future markets, and focused on our strongest sites. We restructured operations to conserve cash and protect our healthiest assets. The posture of the company shifted. Instead of building for growth, we were operating for resilience. Every decision was about survival, not scale.

The Transition: Time to Step Aside

By the end of 2023 it was clear we were not going to build new products or pursue strategic growth for a long time. That helped cement my decision. By 2024 I knew the company didn't need more invention. It needed steady management. After nearly ten years, I decided to step aside. In January I made it official. I raised a $48 million Series C and stayed through a six-month transition to give the company a stable footing for its next phase.

Reflections: What Worked, What Didn't

What I learned building SmartLabs was mixed. Some of it worked. Some of it didn't. Some of it no one could prepare for.

What worked was the proof.
We showed modular labs could function at enterprise scale. We cut response times from days to minutes. We proved you could build once and reconfigure continuously, with facilities shifting half their footprint year to year. We created a new operating model. Not outsourcing. Not pharma-owned labs. Fully operated labs on demand. That pushed the industry forward.

What was difficult was scale.
Running one building is manageable. Running half a million square feet across two coasts is something else. Every weakness in systems, culture, and process shows up. Hypergrowth is possible, but it forces reinvention of operations at every step. What worked at 50,000 square feet broke at 500,000.

We also misjudged things.
We underestimated how long it would take to educate both sides of the market. Customers had to learn to run R&D in a new way. Landlords and investors had to learn we were not another WeWork. That education took years. We thought we were building for five or six therapeutic categories. It turned out to be dozens. That complexity slowed adoption, but it also validated modularity. Constant change was the only model that worked.

There were surprises. The caliber of people who joined us. Hans, who had run the diagnostics divisions globally at Novartis and Bayer. Saquib, who had led infrastructure and facilities worldwide for Boston Scientific. Brian, who had run GE Healthcare's global biomanufacturing solutions business. They left senior global roles to join us because we gave them something large companies could not: the mandate to fix what they knew was broken. That culture of permission to design better systems was as critical as the technology itself.

And there was what no one could prepare for.
The collapse in 2023 was sharper than anything I had seen. IPOs disappeared. Private capital for pre-revenue companies dried up. Almost half our clients were hit. That kind of market freeze is outside anyone's playbook. You adjust, but you do not control the market.

Looking back, there are things I could have done differently. We might have educated faster, or built slower, or picked our bets differently. But I believe we got the big calls right. We invented a new model. We scaled it. We proved it worked for companies that changed the industry.

I am proud of what we built. We gave small teams the same infrastructure as Big Pharma in weeks instead of years. We enabled breakthroughs that reshaped the decade, including the first FDA-approved CRISPR medicine. We created the Lab-as-a-Service category. And I learned more in those ten years than I could have anywhere else.