MEP Infrastructure - Forum Ventures AI Studio

The AI buildout has a skilled trades problem. We want to co-build the fix.

The US needs 300,000 new electricians over the next decade. Electrical work alone accounts for up to 70% of data center construction costs. The bottleneck isn't capital or land. It's qualified technicians. Forum AI Studio is looking for a founder who has run MEP projects and wants to build the solution.

Co-Build With Us →

Experience doesn't scale. Right now it just retires.

The people who know how to commission a 50-megawatt data center cooling system, balance a high-density power distribution panel, or troubleshoot a chiller under load learned it over ten to fifteen years of working alongside someone who knew more than them. That knowledge transfer pipeline is broken. Senior technicians are retiring faster than they can be replaced. The US loses over 400,000 skilled tradespeople per year to retirement alone.

The commercial MEP subcontractors building hyperscale AI infrastructure for Microsoft, Meta, and Google cannot find enough qualified people to staff their projects, and the projects are only getting bigger. Electrical work accounts for 45 to 70% of total data center construction costs. What those teams are working with today: paper checklists, outdated PDFs, and phone calls to the one senior technician who might know the answer.

Three shifts converging at the same time.

The AI buildout is the forcing function

Data center construction spend is projected at $500B+ globally through 2030. Every one of those facilities requires licensed MEP work, and the demand is accelerating faster than any training pipeline can respond to.

The senior workforce is aging out

400,000+ skilled tradespeople retire every year with no replacement pipeline. The knowledge leaving the industry is irreplaceable through traditional apprenticeship alone. The only way to close the gap is to make that knowledge available at scale.

AI can finally close the gap

Voice interfaces, multimodal AI, and AR overlays can now deliver expert-level guidance in a noisy industrial environment in real time. The technology crossed the threshold. The product hasn't been built yet.

The entry point shapes the whole company.

All four are fundable. All four address the same broken market. What matters is which one matches your background.

Path A

AI commissioning co-pilot

Real-time voice and screen guidance that walks a less-experienced technician through complex commissioning sequences, cross-references equipment specs, flags deviations in real time, and escalates to a remote expert when needed.

—Sold to MEP subcontractors per technician seat or per project
—Immediate ROI: measurable reduction in rework and commissioning time
—Data moat builds with every job completed

Path B

Knowledge capture platform

Structured capture of expert technician knowledge through observation, video, and sensor data, turning what your best people know into reusable procedural AI that every technician on every shift can access. The defensible version builds a skills graph that compounds over time.

—Sold to contractors as workforce continuity infrastructure
—Compounds with scale: more technicians, richer model
—Addresses the retirement cliff directly

Path C

AI-native MEP services firm

Don't sell software to contractors. Be the contractor. Build a commercial MEP services company that uses AI to run fewer senior technicians across more projects, competing on speed, consistency, and guaranteed commissioning timelines.

—Faster path to revenue than pure SaaS in this market
—Vertical integration creates a durable moat
—Exits to large MEP firms or general contractors as strategic buyers

Path D

Credentialing and workforce pipeline

Build the pipeline that puts qualified technicians on the job in the first place. AI-accelerated training, assessment, and credentialing that compresses the path from entry-level to job-ready, then places vetted talent directly with the contractors who need them. You own both the supply of skilled labor and the proof that it is qualified.

—Two-sided model: training revenue plus placement and staffing fees
—Credentials become the trusted standard contractors hire against
—Attacks the root supply shortage, not just per-job productivity

A founder who has lived this problem, from any side of it.

Maybe you've run MEP projects and know the commissioning floor firsthand. Maybe you've built workforce or training platforms and understand how skilled trades pipelines actually work. Maybe you've sold software into contractors and know what they'll actually pay for. Whichever background matches the path above, we want to hear from you. You don't need to have done all four, you need to know exactly where this industry is broken and be relentless about fixing it.