What It Takes to Build a Deep-Tech Company
A conversation with Arnaud Legris
Arsenale is building a neo-industrial biomanufacturing company – one that integrates modular infrastructure, advanced sensing, and AI-enabled design to radically transform how biology is deployed at scale.
As Chief Operating Officer, Arnaud Legris is connecting the dots across a complex system, with an evolving role as the company grows.
In this conversation, Arnaud Legris shares insights on:
How his experience at BCG Deep Tech and industrial projects shaped his approach
What it takes to operationalize a deep-tech company from the ground up
Why mastering complexity - not avoiding it - is key to building scalable biomanufacturing
Arnaud, you co-founded Arsenale after leaving BCG, where you were part of its Deep Tech practice. What made you take the leap into building something from the ground up?
At BCG, I had the privilege of working across a range of deep-tech ventures - spanning synthetic biology, AI, quantum. What I saw, again and again, was a paradox: the scientific talent was world-class, but the systems required to scale those technologies industrially simply didn’t exist. The innovation wasn’t being blocked in the lab, – it was being blocked by mindset and structure. People start from a technology which is great in the lab, but are disconnected from the industrial implementation. Worse, by not starting from the underlying problem, their solution becomes just one piece of a larger puzzle – disconnected from the operational context where complexity accumulates, but where real value is captured.
Co-founding Arsenale was an opportunity to move away from the consultant role and return to my engineering roots. I wanted to be part of the growing wave of synthetic biology and actually start building the solution to bring it to scale. It’s a different kind of work. We’re not just solving for performance – we’re building a system that can carry complexity, not strip it away.
You’ve spoken before about the need to build a new Operating System for deep tech. What does that look like at Arsenale?
We use that phrase deliberately. Most companies still try to scale innovation without embracing its complexity - isolated teams unaware of each other’s work and / or missing the big picture, siloed information, long feedback loops. But deep tech operates on a different logic: rapid iteration, convergence of disciplines, continuous learning. That’s the essence of the DBTL cycle - Design, Build, Test, Learn, but company wide. It’s fast, it’s intense, it’s feedback-heavy, and it requires alignment across biology, engineering, data, lab, industrial, operations.
So we’re not just building machines. We’re designing a company that can metabolize complexity. That means flatter structures, real-time and iterative decision loops, deeply integrated teams. It also means being honest about how many different mindsets need to coexist: the builder, the operator, the biologist, the industrial engineer.
And it humbly means rewiring expectations – not assuming someone else has solved a downstream issue, but actively thinking about the impact of your work on others. Are you generating usable data? Are you building for operability? Are you designing with the right skillsets in mind? That’s the shift.
What’s the role of the COO in a company like this? What have been some of the hardest lessons so far?
“Chief Operating Officer” is just a title – you could just as well call it firefighter, gardener, or architect. At its core, the role begins with building the infrastructure and processes that enable teams to execute.
But this is building while running – and from scratch. It demands constant problem-solving, with new challenges surfacing every day, while capturing the lessons that will shape better systems over time. There’s everything to build and no ready-made blueprint.
It’s a deeply architectural task. Not about perfect plans, but about making the organization legible to itself - so it can learn, adapt, and move.
My role is to maintain a transversal view across the company: solving for missing links, wearing multiple hats, and enabling others to focus on what they do best. It means being in the trenches – not optimizing from above, but making sure things get done.
One of the most important lessons learned is that execution is everything. We operate across distinct domains - biology, engineering, data, lab, industrial, operations. Beyond assembling the talent, we have to build the connective tissue that enables real collaboration, grounded in situational awareness.
Looking forward, what’s next for you and for Arsenale?
We’re moving toward stabilizing and learning from the fast growth and development phase. In less than 1 year, we grew from 8 to almost 30 people, we experienced interfaces across teams, we onboarded hundreds of suppliers, two laboratories and industrial infrastructure came out of the ground, we implemented new tools which require training, and much more. The time has come to structure and strengthen our foundations to accelerate.
We’re not just building bioreactors. We’re building a new kind of company. And if we do it right, others will be able to follow - not by copying the surface, but by adopting the underlying principles.
Arnaud co-authored a BCG + Hello Tomorrow report, Nature Co-Design A Revolution in the Making, which explores the specific challenges and breakthroughs required to scale synbio ventures. His current work at Arsenale brings that vision into operational reality.