After about six years working in tech, I took a short contract helping a friend build a cleanroom in Jupiter, Florida. He owns a company that sells cleanrooms and data center infrastructure, and he needed labor for an installation. I had construction experience through my stepdad, who is a superintendent. Working with my buddy, getting paid, and putting that experience to use seemed like an easy decision.
We completed the job in five shifts. The enclosure was intended for biomedical peptide research and production.
What interested me was the business behind it. My friend supplies the systems these facilities need, and he often does the installation himself to reduce overhead and retain more of the profit. His business depends on understanding a technical product well enough to sell it, then taking responsibility for getting it installed.
That gave me a different way to think about where someone with my background could contribute as these industries expand.
The construction behind pharmaceutical growth
Cleanrooms have been around for decades. The growth in pharmaceutical manufacturing, along with demand for computing infrastructure, is giving established construction specialties more attention. Both require facilities that can support expensive equipment and demanding operating conditions. Someone has to turn those requirements into a space that works.
The scale of pharmaceutical investment helps explain why that work matters. In June 2024, Novo Nordisk announced a $4.1 billion manufacturing expansion in Clayton, North Carolina, with plans to add 1.4 million square feet of production space. The company expected to engage about 2,000 external contractors at the project's peak. That is a considerable amount of construction supporting the production of injectable medicines. Novo Nordisk's expansion announcement.
Eli Lilly's February 2025 announcement offered another useful measure. Across four planned U.S. manufacturing sites, it anticipated more than 3,000 permanent positions and nearly 10,000 construction jobs during development. Those were company projections, but they illustrate how much work happens before a facility begins making medicine. Lilly's manufacturing announcement.
Our job in Jupiter was a small enclosure inside an existing building. It served a different purpose and operated at a completely different scale from those projects. Still, working on it made the connection between research and construction much easier for me to understand. A company can have funding, equipment, and a research plan, while still needing a crew to assemble the environment where the work will happen.
What a smaller supplier can offer
The modular construction was particularly interesting. The enclosure used specified panels and connections, with partitions, pass-throughs, and filtration equipment arranged for the space. Prefabricated components gave the installation a defined structure, while leaving plenty of work for the people interpreting the drawings and fitting everything together.
That has a practical business application. A customer working inside an existing building has constraints that a product catalog cannot resolve on its own. Access, layout, and installation all affect whether a proposed system is suitable. A supplier who understands those details can make better recommendations and take more responsibility for delivery.
I could see why my friend stays involved in the labor. He earns more of the project revenue by doing it himself, but he also takes on the time and execution risk. Every hour spent installing is an hour he cannot spend selling another job. The margin has to account for that work.
To me, that is an interesting question for a small company: how long does it make sense for the owner to handle both sales and installation? Doing both gives him direct knowledge of the product and the customer's problems. Growth eventually requires deciding which responsibilities to keep and which to hire for.
Computing has physical requirements too
His other main market, data center infrastructure, raises similar questions about the capacity to deliver. The International Energy Agency reported that global data center electricity demand grew 17% in 2025. Its updated central projection puts consumption at roughly 950 terawatt-hours in 2030, compared with 485 in 2025. The agency also identifies constraints involving electricity supply, grid connections, and equipment manufacturing. IEA's 2026 analysis.
Electricity demand does not tell us how many installation contracts will become available. It does show why the physical requirements of computing deserve attention. Expanding digital services means making decisions about facilities, cooling, power, and the people qualified to deliver them.
Where my experience fits
Having worked in software, I tend to approach a technical problem by reading the documentation, understanding the dependencies, and checking my assumptions. Construction experience makes me consider access, handling, and what someone will actually have to do on site. In Jupiter, both ways of thinking were useful.
I also had plenty to learn. Familiarity with construction and software does not make me a cleanroom specialist. Completing an enclosure is one part of preparing a research or production facility, with electrical work, testing, and operational requirements extending beyond that installation.
But the trip gave me a more specific interest in this work. I want to understand how smaller technical contractors choose their products, price installations, and build reliable crews as customer demand changes. Those decisions determine whether increased investment becomes a sustainable business for the people doing the work.
I went to Jupiter because my friend needed help. I came home interested in a field where my construction experience and years in tech could both be useful, and with a better idea of what I would need to learn next.