What 10 Provisional Patents Tell You About the Future of Contractor Software

A patent is an expensive, time-consuming way of saying "we think this matters enough to defend." Most SaaS companies do not bother. They ship features fast, iterate, and rely on execution to stay ahead. That is often the right call.

Dominus Foundry has taken the other path — 10 provisional patents filed, covering the core innovations in Forge. The portfolio is not a wall of defensive filings. It is a thesis about what commercial contractor software will look like in 2030, and we are betting on that thesis in a way that costs real money up front.

This article walks through the four categories the patents cover, what each one claims, and what that tells you about where the industry is going.

Category 1: Spatial Intelligence

Three patents cover the spatial intelligence stack — how LiDAR capture, AI interpretation, and structural analysis work together.

The first covers the capture pipeline itself: a hybrid AVFoundation-plus-ARKit architecture that fuses camera, LiDAR, and IMU data in a way optimized for construction-specific geometry. The novelty is in how the pipeline handles occlusions (mechanical equipment, parapet walls) that cause standard ARKit scans to degrade.

The second covers automated structural adequacy determination — the system that looks at a 3D roof model, identifies the existing structural system from visual cues (deck type, joist spacing, span estimates), computes the load added by a proposed new roofing system, and flags structures that cannot support the added weight. Nothing like this has existed for commercial roofing before. It is not a replacement for a licensed engineer's stamp, but it catches 80% of the problems early — before a contractor quotes a job they cannot economically perform.

The third covers the interaction model: how scan data flows into measurement reports, CRM proposals, and material takeoffs as a single pipeline. The novel claim is the zero-export workflow — scan data never leaves the system boundary on its way to becoming a quote.

**What this says about the future:** Commercial contractor software is going to have direct measurement as a primary input, not a secondary integration. The winners will be tools that treat LiDAR data as a first-class citizen in the data model, not tools that import an EagleView PDF and call it done.

Category 2: Communication Intelligence

Three patents cover the Torch communication module.

The first covers compliance-preserving call analysis — how a pipeline can analyze a call for TCPA compliance, customer sentiment, and negotiation signals without retaining raw audio beyond what is legally required. The novelty is in the data retention schema: which derived signals are kept permanently (sentiment scores, compliance flags) versus which are purged on a rolling basis (raw audio, transcripts).

The second covers multi-party conference transcription with speaker identification for contractor sales calls specifically. Standard transcription tools (Otter, Fireflies) struggle with three-way calls between a rep, a customer, and a spouse or partner. The patent covers techniques that improve speaker separation for those specific scenarios.

The third covers AI-driven negotiation coaching — a system that analyzes an in-progress call in real time and provides the rep with talking points, pricing guidance, and compliance warnings. This one is still largely in R&D, but the IP position is filed.

**What this says about the future:** Every sales call is going to be recorded, analyzed, and used for both compliance and coaching. Reps who do not get this support will be at a disadvantage. The industry winners will integrate communication intelligence as a core function, not a bolt-on.

Category 3: Compliance Automation

Two patents cover the Compliance Engine.

The first covers automated permit routing — a system that takes a job site address, identifies the relevant AHJ (authority having jurisdiction), determines the permit types required for the scope of work, and either submits applications electronically or generates the correct forms for manual submission. The novelty is in the AHJ routing model, which covers 2,400+ jurisdictions in the US with jurisdiction-specific variations in form content, submission channels, and fee structures.

The second covers code compliance tracking across a job lifecycle — a system that identifies applicable codes (IBC, IRC, NFPA 72, NEC) based on project scope and maintains a living compliance record as the job progresses. The novel claim is in the automated evidence collection — photo documentation, inspection records, material certifications are automatically linked to the specific code requirements they satisfy.

**What this says about the future:** Compliance is going to shift from a manual burden to an automated baseline. Contractors who continue to manage permits and inspections on paper will lose to contractors who handle it in software. This is an area where small operators can leapfrog larger ones, because the software does the work a compliance manager would otherwise handle.

Category 4: Negotiation and Deal Systems

Two patents cover how Forge handles the actual negotiation of a commercial job.

The first covers an AI-driven proposal generation system that composes pricing based on historical win rates at given price points. The system learns which price points close which kinds of customers and biases proposal pricing toward optimized win probability. The novelty is in the learning architecture: the system avoids racing to the bottom on price by explicitly modeling customer lifetime value, not just close rate.

The second covers a multi-round negotiation support system — tracking pricing revisions, counter-offers, and scope changes across the life of a deal. Current CRMs treat pricing as a static field; this one treats it as a timeline.

**What this says about the future:** Commercial sales is going to look more like enterprise software sales — multi-round negotiations, managed pricing, optimized close rates — than residential retail. The contractors who adopt that model early will win larger deals at better margins.

Why File Patents as a Startup

A common startup position is "we will move fast and worry about IP later." That position is sometimes right. It is wrong when the core innovation is genuinely novel and when you expect larger incumbents to enter the category.

Both of those conditions apply to commercial contractor software. The incumbents — ServiceTitan, Procore, Autodesk — are not going to ignore this market forever. When they enter, the IP position matters. A defensible portfolio makes acquisition talks easier, makes partnerships more favorable, and reduces the risk of being out-shipped by a larger competitor.

The cost of provisional patents is manageable (low thousands of dollars per filing). The value is optionality. If the category grows and incumbents enter, the patents matter. If the category stays niche, the patents are a manageable cost. The expected value is positive.

What This Means for Contractors Buying Software

Most contractors do not care about patents directly. You care about the software working. But the patent portfolio is a signal about seriousness. Companies that file real patents are committing to a long-term technical direction. Companies that do not are keeping their options open — which is fine, but it is not the same commitment.

When you are evaluating contractor software, ask vendors about their IP position. Not because patents guarantee anything but because the answer tells you how they think about the technology. Vendors who treat their product as genuine IP tend to invest more in the underlying architecture. Vendors who treat their product as a configurable dashboard tend to invest more in the marketing.

Both are legitimate business models. They just produce different kinds of software.

FAQ

**What is a provisional patent?** A provisional patent application establishes a priority filing date for an invention, giving the inventor 12 months to file a full utility patent application while claiming the original date. It is a less expensive way to lock in IP priority while doing the work of turning the invention into a full patent.

**Are 10 patents a lot for a startup?** For a Series A-stage startup, 10 provisional filings is more than typical but not unusual in IP-heavy categories. The comparable point is less the count and more the scope — patents that cover core architecture tend to have more strategic value than patents that cover narrow features.

**Does Forge license any of this IP?** Currently, no. The portfolio is used internally. If the industry evolves toward licensing (as happened with CAD in the 1990s), that is possible, but it is not the current plan.

**When do these become issued utility patents?** The provisional filings convert to full utility applications within 12 months of the original filing date. Issuance timelines are 2-4 years depending on the USPTO examination queue.

**Can a contractor use a competitor's LiDAR tool without infringing?** Generally yes — provisional patents are not enforceable until they issue as utility patents, and even then, infringement is a specific legal question. The patents cover specific implementation details, not the general concept of using LiDAR for roof measurement.