Reach a highly engaged audience of AEC innovation leaders, executive teams, and industry titans. Check out our media kit.

A narrative has been spreading through AEC tech circles. A well-funded competitor announced free AI contract review. Pre-signature risk analysis, post-signature compliance, delivered in under 30 minutes. The message from the market was pointed: some AI capabilities should be table stakes now.

The implication was clear. Document Crunch's core product had been commoditised overnight.

Josh Levy, CEO of Document Crunch, saw it. We sat down with him to understand his response, and what it reveals about trust, defensibility, and what the industry should be paying attention to.

A competitor launched free AI contract review and the market cried “commoditised.”

Josh Levy says that’s the wrong frame.

If you’re selling a feature, free wins.

If you’re selling trust in a high-stakes industry, it’s a different game.

Free is arriving in AEC tech with serious capital behind it. Contract review became the first product to be declared a commodity in the narrative. That declaration is worth examining carefully, because the stakes of getting contract review wrong are not abstract.

The Arcadis 2025 Global Construction Disputes Report puts the average dispute value at $60.1M, up 40% year-on-year, with an average resolution time of 12.5 months. For the third consecutive year, the leading cause of those disputes was errors and omissions in contract documents. 

Construction attorney Michael Vardaro, speaking in a previous Bricks & Bytes conversation, described the same pattern from the courtroom: scope conflicts that could have been caught pre-signature instead compound across the project lifecycle and become the disputes that drain everyone.

When a free tool enters a category carrying that kind of consequence, the right question is whether the product is actually doing the same job.

Levy's first argument is definitional. The label "contract review" is being applied to two different products, and treating them as equivalent is a category error.

Document Crunch has spent seven years processing construction documents across 17 document types, building proprietary benchmarking infrastructure, and embedding accountability into every output. In construction, the output of contract review is a paper trail. Who flagged what? What was missed? What was signed off on? As Levy frames it, a free tool has no institutional skin in the game. The accountability sits elsewhere, usually with the GC's legal team that relied on it.

Levy's second argument runs deeper than product features.

"You cannot vibe code your way to trust in this industry," he told us.

In his view, trust in risk management is built through authentic relationships, deep empathy with how construction projects actually run, and a track record of accuracy when the consequences are real. It accumulates over years of being right, being corrected, and being trusted again.

This is where Levy invokes Roosevelt's words at the Sorbonne in 1910. The credit belongs to the person actually in the arena, whose face is marked by dust and sweat and blood. Levy has been in that arena, shipping to GC legal teams, iterating through failures, and earning the institutional relationships that make Document Crunch's outputs defensible when a dispute lands.

Andrej Karpathy coined the term "vibe coding" in February 2025, now Collins Dictionary's Word of the Year. The concept describes using AI to generate outputs rapidly and intuitively, without the rigor of structured engineering discipline. The surface looks right. The underlying risk is harder to see.

The Veracode 2025 GenAI Code Security Report tested more than 100 LLMs across 80 tasks and found that 45% of AI-generated code contains security vulnerabilities, with performance not improving despite advances in the underlying models. Levy's point is that replicating the output surface is different from replicating the accountability, the benchmarking rigor, and the institutional relationships that make the output defensible when a $60M dispute arrives.

Levy does not frame this as a product war. His position is a provocation to the market.

"If someone can vibe code their way to putting Document Crunch out of business, so be it. The market will speak and I'll leave the arena."

He is betting that does not happen. Seven years of commercial traction, a $21.5M Series B with contractor investors as LPs, and a platform expanding across 17 document types suggest his confidence is grounded in something beyond posture.

The confusion created at the category level is the real competitive threat. Free contract review and accountable contract review share a name but carry different risk profiles. A GC relying on a free AI tool for a $150M project absorbs the liability. A GC relying on Document Crunch has a documented, benchmarked audit trail. They will be treated as equivalents unless the distinction is actively defended.

The company that defines the evaluation criteria for what good contract review looks like will carry a structural advantage. Harvey AI understood this early. Now valued at $8B, it publishes benchmarks and actively shapes how the legal AI market assesses quality. Document Crunch has the internal infrastructure to do the same. The company that defines the standard tends to own the narrative.

GTM Lesson: Allowing a competitor to define what your category means cedes the framing war, regardless of how the products compare on technical merit. Category definition is a commercial asset.

Come find me in the arena. Build the trust. Earn the customer. That is Levy's challenge to the market.

For anyone watching this moment play out in AEC tech, the underlying question is about what you are actually selling. If it is a discrete feature, free will eventually win. If it is a trust position built over years in a high-stakes industry, the calculus looks different.

Defensible contract review is built through years of accountability. Levy's bet is that the market will recognize the difference. Watch the video.

Has a competitor or a free tool ever threatened to one-shot your core product? What happened, and how did you respond? Reply and tell us. We read every one.

Puzzler: The "Man in the Arena" speech was delivered at which institution, and in which year? Reply with your answer. We will shout out the first correct respondent in the next issue.