Maegan Spivey’s journey from construction project administrator to Product Manager at Document Crunch gives her unique insight into building AI products for the sceptical construction industry. In this conversation, we explore how domain expertise, radical transparency, and human amplification enabled Document Crunch to achieve an 80% reduction in contract review time while earning trust in an industry burned by many solutions.
About Maegan Spivey
Maegan Spivey is an Associate Product Manager at Document Crunch, an AI Risk Reduction Platform for the construction industry. She brings experience across retail, multi-family, and energy construction markets and has worked for both general contractors and subcontractors. Her background spans project administration, accounting, and contract management roles. Today, Maegan applies that experience to help solve industry-wide challenges with construction contract documents.
Introduction
“It’s easy to be passionate about this work when you’ve lived its pain points.” Maegan Spivey
Construction has created documentation so complex, so voluminous, so impossible to comprehend that we now need artificial intelligence to navigate it. Each addition seemed justified in isolation, liability concerns, regulatory requirements, risk management, insurance mandates, and lessons from past disputes. Rational decisions that collectively produced documents nobody can comprehensively understand within tight project timeframes.
Maegan laughs. The pitch is simple: Do you like reading contracts? The answer is always the same. No. Not the business owners forced to review agreements. Not the small companies without counsel. Not the project teams drowning in specifications. She was able to build a career in contracts because she was the only one who enjoyed reading them.
This universal pain reveals Document Crunch’s core insight: they’ve built an AI product that achieves an 80% reduction in contract review time, not by doing something radically new, but by making a longstanding professional standard, thorough contract review, achievable again.
But here’s what makes their story particularly instructive: they’re succeeding in an industry burned by overcomplicated technology, selling AI to professionals sceptical of both new software and artificial intelligence, solving a problem the industry created for itself. Three interconnected insights emerge from their path: solve genuinely universal pain points, leverage domain expertise at every level, and build trust through radical transparency and human amplification.
The pain point everyone recognises
As construction tech expert Tannis Liviniuk has noted, the biggest driver of rework in construction isn’t poor craftsmanship; it’s issues in the contract documents. “There’s so much information, you can’t humanly go through it without sacrificing other work”, Maegan explains. We’re tearing down and rebuilding work because nobody could find the relevant information in time, because conflicts between the contract documents weren’t caught, and because compliance requirements got lost in the volume.
This is what the tool addresses: restoring the ability to do a comprehensive review at all. AI doesn’t enable something new; it makes what should be standard practice actually possible within project and human constraints.
The domain expert advantage
What makes Maegan’s perspective valuable isn’t just knowing contracts are painful; it’s understanding the precise ways complexity accumulated, the defensible reasons for each addition, and the collective irrationality of the result.
Her path to product management wasn’t typical. She started as a project administrator, fell into contract management almost accidentally, and began sharing contract concepts on LinkedIn. That content caught CEO Josh’s attention at Document Crunch in 2023 and she moved from construction to technology. About a year ago, she audited her tasks to find what she enjoyed most: working with customers, testing features, connecting feedback to improvements, and plugged the list into Perplexity, ending with, “What job is this?” “It went to product manager.”
This transition from doing the work to building tools for the work creates advantages hard to replicate. The Document Crunch team includes people who held the roles they’re now building for. “We’re in the product, we’re using it,” Maegan says.
When you’ve been responsible for contract compliance on a live project, you viscerally understand why “just trust the AI” doesn’t work. You remember disputes and damaged relationships. Accuracy isn’t nice-to-have; it’s the difference between protection and liability.
But domain expertise creates its own challenge:
“The way you did it is not the way every other company has done it. So keeping an open mind of what are the other ways, is there a better way than what you experienced yourself?”
The tension between deeply understanding the problem and avoiding building only for your own workflow defines the practitioner-turned-product-builder challenge. Small construction companies and large ones vary significantly. You must keep your own process loosely held.
The solution: combining internal expertise with external feedback. Some evolution comes from the team’s professional knowledge; they read contracts, recognise patterns, and see logical next steps. Other expansion comes from customer conversations about related pain points.
This is where a critical role in AEC innovation emerges: subject matter experts with entrepreneurial mindsets bridge the gap between understanding industry pain and building solutions that work. Not just people who understand the industry, but people who’ve done the actual work, felt the actual frustrations, developed instincts from repeated exposure to real problems, and then turned that experience toward building solutions.
The flywheel: customer stories drive product excellence
Maegan describes a powerful mechanism at the heart of product development: a flywheel in which user narratives educate the team, better understanding improves the product, and an improved product creates more success stories that further illuminate the problem.
“I love seeing the team, who haven’t worked in construction, being able to hear what this means to customers,” she explains. “It lets customers bring the problem of disputes to life for people who are supporting the construction industry.”
This connects to a broader truth about product development that Sol Amore, Product Manager at Autodesk, articulated in a previous conversation: successful products emerge from understanding “the spirit of the ask,” not just what customers say they want, but the truth behind their requests.
This flywheel solves a critical challenge in AEC product development: how to build for an industry when your team hasn’t lived in that industry. Practitioner knowledge provides the foundation, but customer feedback prevents it from becoming myopic.
The stories do more than inform features; they create shared language across departments. When software teams hear how document conflicts destroyed a project relationship, disputes shift from abstract legal proceedings to relationship-destroying events. The mission becomes tangible rather than aspirational.
“We’re talking to customers, but customers give us stories and we’re able to make the problems really come alive,” Maegan says. The flywheel works because it centres on customer problems, not product features, and talks about what customers are already talking about, not what the company wants to talk about.
This mechanism distinguishes products that genuinely serve users from those that merely implement features. The flywheel ensures that firsthand understanding doesn’t calcify into assumptions, that customer pain stays visceral rather than theoretical, and that the entire team, not just customer-facing roles, understands what success actually means.
For other AEC product builders, this suggests a specific practice: systematically bringing real-world examples into engineering conversations, not as requirements but as context. Make the pain tangible. Let customers educate the team. Build the flywheel deliberately.
Building trust in a burned industry
Construction professionals already distrusted software before AI entered the conversation. Maegan witnessed this firsthand. At one company, new software that was “supposed to make everyone’s life easier” actually tripled her workload.
This scepticism isn’t irrational; it’s learned. As Harvard Business School professor Frances Frei notes:
“It’s not that people don’t like change, it’s that usually change makes things worse.”
AI tools face an even higher bar, as they introduce opacity in an industry where transparency determines trust. Add media narratives about hallucinations and job replacement, and you have scepticism layered on top of all the others.
The trust-building strategy reflects direct understanding of why professionals are sceptical:
Radical transparency. When the AI identifies risks or answers queries, it shows exactly where in the document it found information. “We’re not trying to remove your ability to check an answer,” Maegan explains. “We want to make it easy so that you can do that.”
The transparency also extends to limitations. For example, English-based construction contracts are their primary focus. They won’t start claiming to provide those same services for European language-based contracts. “If we don’t have people in-house who can check Italian answers and validate them, that’s not our wheelhouse right now.”
Human amplification, not replacement. AI doesn’t eliminate judgment; it makes exercising judgment faster and more comprehensive. It can surface more relevant information than humans could in the same timeframe, then let humans make final judgment calls.
Quality obsession. “If we’re only giving 50% accuracy on anything, we’ve made more work for you because you have to go figure out what 50% is missing.” Testing involves people who actually did the work, such as attorneys or contract professionals, checking against the standard of what good contract review looks like in practice.
Testing also allows customers to focus on answer quality, rather than the noise and hype around the latest models. There has to be a benefit to customers in making changes. The competitive advantage comes from the quality assurance layer around the AI, not the AI itself.
Human relationships that acknowledge reality. “When I think about introducing somebody to AI, it’s: come and see, come and try it for yourself, and let’s work through it a little bit together.” The customer-facing team is, in Maegan’s words, “really kind-hearted, genuine people.” They build the relationships that can help introduce AI in an approachable way.
Why starting narrow enables expanding wide
The team started extremely narrow: speeding up construction contract reviews. Not contracts generally. Not document management broadly. Just reading through a construction contract and identifying common risks quickly enough to be useful.
This narrow focus reflects what Karri Saarinen from Linear observes about effective AI products: “What is the problem someone has today?” Rather than building broad AI tools that can “do anything for anyone,” the value comes from solving specific, high-impact problems where AI can provide immediate value.
But contracts don’t exist in isolation. Over the past year, they’ve expanded into specifications review, compliance documentation, notice generation, and project team playbooks. The expansion follows a deliberate logic grounded in both domain expertise and customer feedback.
“When we think about disputes, it’s never only just one clause in a contract,” Maegan explains. “There are so many risks of disputes in other areas.” Specifications may include requirements that conflict with contract terms. Compliance obligations that must be met.
This expansion strategy serves the recently updated mission: “zero disputes and stronger relationships in the construction industry.” The vision parallels the shift in construction safety culture, where “zero incidents” moved from aspirational to achievable.
“Similar to the safety movement, we don’t want any more disputes,” Maegan explains. “It’s kind of cyclical when you have a stronger relationship, you’re more resilient, so you’re less likely to have disputes. And fewer disputes means you also have a stronger relationship.”
The mission serves practical purposes beyond inspiration; it helps prioritise among construction’s overwhelming pain points. “There are a lot of pain points in construction. All these solutions can be developed right now with rapid prototyping and AI capabilities. The mission and the vision help us, from a product perspective, think about what comes next.”
Each expansion connects logically to the core problem while addressing real workflow needs. It’s not feature creep, it’s recognition that the documentation crisis extends across the entire project lifecycle.
The strategy also reveals something about building in AEC: start narrow, achieve genuine excellence, then expand to adjacent problems where initial trust transfers. They didn’t jump from contract review to project management; they moved to compliance and specifications, where the same core capability (extracting insight from dense, legally-binding text) applies and where the same frustrated customers already trust them.
The pattern that emerges
The path Document Crunch has taken reveals a fundamental truth about successful product development: focusing on problems rather than solutions dramatically increases your likelihood of success. As one product leader put it, if you focus on the solution first, “you are going to build a solution and then you will be going to search for a problem that this solution creates value for. And so your journey is going to be way longer.”
But when you can say “I am solving this problem for these types of users,” you have a North Star guiding every decision. This is exactly what Josh Levy and his cofounders achieved: now others continue to build on it. They identified a universal pain point, understood the users deeply, and built from there.
Their path reveals three elements that don’t just coexist—they form a reinforcing system:
Solve genuinely universal pain. “Nobody likes reading contracts” is brutally simple, but that simplicity makes it powerful. Everyone in construction recognises documentation overload. The opportunity comes from solving problems the industry created for itself, where attempted solutions compounded into new problems.
Leverage domain expertise at every level. Having people who did the actual work shapes product decisions, quality standards, trust-building strategies, and expansion priorities. It’s the instincts and judgment from repeated exposure to real problems. This expertise prevents the creation of features nobody needs and helps articulate problems in ways that resonate because you’ve lived them.
Build trust through transparency and human amplification. In an industry burned by bad technology, now facing AI scepticism layered on software fatigue, trust must be earned through radical transparency, positioning AI as amplification rather than replacement, obsessive quality assurance, and relationship-building that acknowledges legitimate anxieties.
Universal pain attracts practitioners who have lived it, giving Document Crunch the domain expertise to build solutions that genuinely work. Working solutions build trust in a sceptical market, generating customer stories that educate the entire team about construction’s realities. An educated team builds better products, which solve adjacent pain points, attracting more domain experts who recognise problems worth solving. This cycle explains why Document Crunch succeeds; they built a system where domain knowledge, customer trust, and product excellence compound rather than compete.