Build's initial offering is selling labor. Not software, not seats: outcomes. This comes with its own complexities, but it allows us to tackle institutional projects with a precision that pure software cannot match, pairing agentic AI with human domain expertise to deliver the work itself.
Our long-term plan extends well beyond the institutional market, to development firms of every size, in every geography, running every asset class. The overarching purpose of Build is to help shift the built world from a fragmented, manually executed industry to one where development intelligence operates at software speed, and where the quality of that intelligence compounds over time.
Critical to getting there is a product that works without compromise at the hardest end of the market first, which is why Build's earliest partners have been everyone from Fortune 500 companies to governments: institutions running some of the most complex development programs in the world. What makes institutional workflows tractable is precisely what makes them large: they are standardised, numbers-driven, and accountable to process in a way that smaller, less formal projects are not.
The honest reality is this will not change the built world on its own, and not quickly. But that misses the point, unless you understand what we are actually building towards.
Almost any new technology initially has high unit cost before it can be optimized and scaled. This is no less true for agentic AI applied to development. The strategy of Build is to enter at the highest end of the market, with clients who have the most complex workflows, the highest data quality requirements, and the institutional credibility to prove out the technology at scale, and then move downstream as fast as possible with each successive layer.
Without giving too much away: the second layer is the mid-market. Growing development firms and asset managers running the same workflows as the institutional players but with a fraction of the resources, teams of five doing the work that used to require fifty. The third layer is wider still: every developer, project team, and capital allocator building at any scale, in any geography, who has ever been constrained by what their organization could physically process.
Build is designed to operate like a software company. To the market, we are a services firm. But the operation is the product, and our margin expectations reflect that. All momentum, including capital, data, credibility, and relationships, goes back into expanding coverage and capability. When a global alternative asset manager uses Build to screen a portfolio of industrial sites in Germany, they are, in a small way, helping build the product that eventually serves the developer acquiring a single site in Ohio. Our ontology makes the ambition simple to state: any project, any typology, anywhere.
The throughput problem
The workflows that move institutional development today take weeks. Most of them do not need to.
| Workflow | Industry standard | Build |
|---|---|---|
| Deal Screening (including underwriting) | 2–4 weeks | 4–12 hours |
| Desktop Due Diligence | 3–6 weeks | 8–24 hours |
| Test Fits | 2–3 weeks | 48–72 hours |
This is not a marginal improvement. It is a structural shift in what a development team can process, decide on, and act on in a given period of time.
Right now, one person can manage roughly 10 five-figure workflows simultaneously. That is the current ceiling, and it is already better than anything the industry has produced. But it is not the destination.
The destination is one person threading 1,000 workflows a week. Not managing them. Not reviewing them. Threading them: setting direction, catching exceptions, compounding revenue. The agents handle everything else.
Each step in the table below represents an order-of-magnitude shift in what one person can manage. The numbers are directional. What changes at each threshold is the class of problem being solved, not a roadmap date.
| Ceiling | Workflows / person / week | What breaks it |
|---|---|---|
| Before | 0.2 | Human analyst: weeks per workflow, manual data gathering, single geography |
| Now | 10 | Agent led, every output verified by a Build domain expert |
| Next | 100 | Verifiability: the system proves outputs are correct; humans audit by exception, not by line |
| Then | 500 | Interpretive layer: agents explain reasoning in domain language; a fund manager can push back without reading a log |
| Horizon | 1,000+ | Full primitive coverage: every step in the building lifecycle decomposed, composable, self-correcting |
A data center developer signed with Build earlier in the year and within a month of partnership they brought three sites to an acquisition stage. The intelligence from our platform gave their team negotiating leverage that was not available to them before: constraint analysis, power feasibility, and site comparison at a speed and depth that changed what they could credibly say across the table. Hundreds of millions of dollars of projects. Some of them may not have happened otherwise. This is day zero.
Each milestone is not the same thing running faster. Each one removes a different class of human dependency entirely. As the models improve, the glue strengthens: the Build ontology becomes more proprietary, more workflows are covered, execution gets faster, interpretation gets sharper. And the ability to take on the liability of a project, to be the partner a developer genuinely trusts with a consequential decision, is the defining factor. That trust is the brand of Build.
The workflows that are complex today will be commoditized. That is not a risk. It is the plan. As throughput scales, the cost of Build's output falls, and we will lead that process ourselves, passing the savings directly to our partners. Just as foundation model prices have dropped continuously, so will the cost of what Build produces. What keeps the company at the frontier is not defending yesterday's complexity. It is always working at the edge of what has not yet been solved: the next class of workflow, the next geography, the next asset class, the next modality. The commoditization of the work we do today is baked into the plan.
The organizations that partner with Build earliest are not simply getting better outcomes with the work they do through us. They are accumulating proprietary data on their own decision patterns, building integrations that deepen with every run, and creating a compounding gap between what they can do and what any late entrant can replicate.
The organizations that are in now are not buying a faster version of what they already have. They are running their development at the speed of technology, with a partner whose value grows with every project, every data point, every workflow completed.
Ten years. One observation.
Before Build, I spent years training and practicing architecture, working on infrastructure projects at the scale that shapes cities: airport terminals, cultural institutions, urban districts. The observation that stayed with me was simple. The decisions that produced buildings and places that are genuinely excellent were made by teams who had the time and information to think carefully. That time and information came from working with the 1% of the most wealthy and well-resourced partners in the world. That leaves the other 99%.
Build's commercial strategy, enterprise first, downstream second, wider still, is the mechanism for scaling the product. To tackle the other 99%.
The built world shapes everything. Where people live, how cities function, the energy infrastructure that powers economies, the industrial fabric that employs them. It is, by any measure, the most consequential industry in the world. It is also the one that has changed slowest, not because the people are resistant to change, but because the work has been genuinely hard to systematize and the tools available have genuinely not been good enough.
What access to the world's most significant development pipelines creates, over time, is leverage. Not to impose an agenda, but to raise the baseline. Better-informed decisions produce better projects. Better projects, compounded across thousands of developments over a decade, change what the built world looks like.
So the strategy, put simply:
- Partner with the world's largest institutions to accelerate the world's biggest projects
- Use that capital, data, and credibility to serve mid-market development at scale
- Use that reach to go wider still: any developer, any geography, any asset class
- While doing the above, raise the quality of the decisions being made, and therefore the quality of what gets built
Keep this a secret.