The former CIO of SpaceX and Broadcom explains how to balance process with innovation while building the 'digital nervous system' of a modern manufacturing giant.
The Right Process at the Right Time
In the trajectory of a growing company, the greatest danger is often the premature imposition of structure. When a team is just ten people, it functions as a simple network of smart individuals talking to each other; process is almost entirely unnecessary. As you reach 100 people, you need a light touch of formalization, but the real shift happens at the 1,000-employee mark. This is where many organizations stumble by over-indexing on systems that constrain growth rather than enabling it.
The philosophy I’ve found most effective is 'smart people, stupid systems.' You must assume your people are capable of doing the right thing in the moment, but you build systems that track their actions and hold them accountable. By the time a company reaches 10,000 employees, the challenge shifts again. At that scale, no one understands how the whole business works anymore. The role of leadership and information technology is to synthesize disparate data points and present them to top-level decision-makers so they can see the entire organism operating as one.
Building the Digital Nervous System
When I joined SpaceX, the company was at a critical inflection point. They were building massive metal objects in downtown Los Angeles—an unlikely feat—and producing about one booster per year. Elon Musk’s vision wasn't just to build rockets, but to build the 'digital nervous system' for a 21st-century rocket company. We chose to build a custom platform, Warp Drive, rather than relying on off-the-shelf software. For the first two years, I wasn't sure if this was the smartest or the stupidest thing we had ever done.
By the third year, the brilliance of the approach became clear. We had created a tightly coupled system where a design change made by an engineer was immediately reflected in the projected operational costs. This transparency meant that no one was surprised by a $5 million implementation cost six months down the line; the impact was apparent the moment the change was proposed. This custom infrastructure became a primary differentiator, allowing us to scale to 40 boosters a year and 9,000 employees across multiple locations without the information silos that typically plague aerospace.
The Culture of the Responsible Engineer
Beyond the software, the 'magic' of the SpaceX model lies in first principles thinking and a radical approach to accountability. We operated under the concept of the 'Responsible Engineer.' In most large organizations, a project is a series of handoffs: one person writes requirements, another designs, another builds, and another tests. At SpaceX, one person owned their part from the initial requirements all the way through to post-flight analysis. There were no handoffs and, consequently, no finger-pointing.
This level of ownership creates a unique intensity. People worked twelve-hour days not because they were forced to, but because they were deeply committed to the mission and didn't want to let their colleagues down. This culture was balanced by leadership dynamics that are often misunderstood. While Elon provided the aggressive, focused drive, Gwynne Shotwell acted as the essential glue. Much like the partnership of Henry Nicholas and Henry Samueli at Broadcom, Gwynne was able to translate that high-pressure drive into a sustainable forward motion, ensuring the company didn't burn out as it scaled.
Scaling in the Age of AI
The playbook for scaling is changing rapidly due to AI and robotics. At SpaceX, I managed a team of 175 people to build our core platforms. Today, at Senra Systems, I am working with a team of six to accomplish nearly the same goals. AI allows us to digest data, identify trends, and automate mundane tasks with a fraction of the headcount previously required. We are no longer just throwing more people at a problem; we are throwing highly talented people equipped with tools that eliminate the boring, repetitive parts of the job.
We are currently applying these lessons to the cable harness industry—a massive, fragmented, and largely unautomated business that serves as the backbone for everything from autonomous vehicles to high-tech medical equipment. Most engineers treat harnesses as an afterthought, but by applying the SpaceX playbook—defining the process, making it repeatable, and then replicating it across locations—we can transform a dormant market. It usually takes three versions of a design to get it right; once we have our third facility running, we will have a 'rinse and repeat' model that can fundamentally change the speed and quality of modern manufacturing.