Belidor: the intelligence layer for construction
David Peterson

Today, we’re incredibly excited to announce our investment in Belidor, alongside our friends at Max Ventures. I wanted to share a bit of background about the opportunity Belidor is going after and what got us so excited about what Marc and Baptiste are building.

Construction is becoming a strategic bottleneck for the West. Housing shortages. Infrastructure decay. The inability to build data centers fast enough. The problem is that building takes too long and costs too much. And if you look closely, the bottleneck often isn't on the job site. It's in the office.

Preconstruction, the stage before breaking ground, is particularly rife with busywork. Estimators spend hours every week manually comparing subcontractor bids, copying numbers from PDF to Excel, and hunting through inclusions and exclusions to understand who’s actually providing what. At large GCs, estimators devote fully 25% of their time to this process of rote data entry, instead of actually winning work or moving projects forward. And despite billions invested in construction tech, no one has automated this process in a way estimators trust.

Belidor starts here. Their first product levels bids 100x faster than a human can, pulling data directly from messy PDFs, spreadsheets, emails, or even text message threads, surfacing scope gaps, and linking every conclusion back to the source document.

Belidor is not replacing estimators. It's augmenting them, freeing them to focus on negotiation and judgment, not data entry. And crucially, rather than creating a whole new platform that GCs need to adopt top-down, Belidor mirrors how estimators already work, enabling estimators to adopt the product seamlessly, bottoms-up.

That's the wedge: a painfully specific task that AI can solve remarkably well, in a market where every minute saved compounds across hundreds of projects and billions of dollars in investment.

But the wedge is also the foundation. Bid-leveling is the gateway to the most valuable dataset in construction.

Today, the details of every bid - that is, the true pricing, the scopes, the subs, the inclusions and exclusions - live on individual desktops, in disconnected Excel files, or trapped in email threads. No existing software vendor (not even the major incumbents) has structured access to it. Even most GCs themselves can’t query the historical bids they’ve captured in a consistent way.

Once normalized, that dataset unlocks an entire stack of next-generation, AI-powered construction workflows:

• Scope generation and bid package assembly. By processing thousands of bids, Belidor will learn the relationship between specifications, plans, and actual trade-specific scopes, enabling them to automatically generate scope docs, by trade, from plans.

• Intelligent subcontractor matching. Belidor learns which subcontractors respond quickly, bid competitively, understand scope correctly, and ultimately win and perform well. Combined with project characteristics, Belidor will be able to predict which subs are genuinely well-matched for each job. Less spam, better signal for everyone.

• Automated takeoffs grounded in real data. The hardest problem in preconstruction becomes solvable when you understand how each trade actually scopes their work, what quantities get bid, and which details matter. Plans in, quantities out - in seconds, not days.

• Market intelligence and benchmarking. Real-time pricing across regions and trades. Performance data. Cost trend analysis. The kind of intelligence GCs have always wanted but could never access consistently.

Each new workflow compounds the moat. And each workflow is only possible because of the wedge.

As is often the case, though, the real unlock here is the team. Belidor's founders, Marc and Baptiste, have deep experience in both AI and product design for old-school industries. They're not trying to force-fit a general-purpose AI into construction, and they’re not building a “perfect” product from afar. They're getting on planes, sitting with GCs daily, and purpose-building for how the industry actually works.

Long-term, Belidor will become the intelligence layer for the entire built world. Upstream: architectural designs optimized for constructability and cost. Downstream: intelligent procurement that automatically orders materials at optimal prices and lead times. And the end state: "We need a hospital" becomes a complete design, accurate pricing, qualified subcontractors, ordered materials, and an executable schedule within days or weeks, not months or years.

That's how you create infrastructure abundance. That's how you remove civilization's bottleneck.

But it starts with earning trust, one bid at a time.

The AI-Native Enterprise Playbook
Ten real-time observations on a rapidly evolving playing field

No More Painting by Numbers
It’s the end of the “SaaS playbook.

The Age of Artisanal Software May Finally be Over
Every wave of technological innovation has been catalyzed by the cost of something expensive trending to zero. Now that’s happening to software.

Founders as Experiment Designers
David on why founders should run everything as an experiment.

ENTERPRISE/TECH NEWS

Cold water. In a 2.5-hour podcast, Dwarkesh Patel interviewed Andrej Karpathy on the trajectory of AI. It’s a very thoughtful discussion, and it goes deep into why Andrej does not believe that AGI is likely within the next decade. Overall, it’s a great deep dive into how one of the brightest minds of the era is thinking about AI. Here’s a great example: “I have a hard time differentiating where AI begins and stops because I see AI as fundamentally an extension of computing in a pretty fundamental way. I see a continuum of this recursive self-improvement or speeding up programmers all the way from the beginning: code editors, syntax highlighting, or checking even of the types, like data type checking—all these tools that we’ve built for each other. Even search engines. Why aren’t search engines part of AI? Ranking is AI. At some point, Google, even early on, was thinking of themselves as an AI company doing Google Search engine, which is totally fair. I see it as a lot more of a continuum than other people do, and it’s hard for me to draw the line. I feel like we’re now getting a much better autocomplete, and now we’re also getting some agents which are these loopy things, but they go off-rails sometimes. What’s going on is that the human is progressively doing a bit less and less of the low-level stuff. We’re not writing the assembly code because we have compilers. Compilers will take my high-level language in C and write the assembly code. We’re abstracting ourselves very, very slowly. There’s this what I call “autonomy slider,” where more and more stuff is automated—of the stuff that can be automated at any point in time—and we’re doing a bit less and less and raising ourselves in the layer of abstraction over the automation.”

Where has the creative gone? Hiten Shah wrote a thoughtful reflection on why the early years of the Apple app store were so magical. He suggests that current moment of vibe coding might represent a new momentum of playful creativity that might inject some new energy into the world of software. “Products like Wabi, Sora, Lovable, Notion, Dia, Bolt, and v0 mark a turning point. They prioritize adaptability, play, and personal taste instead of uniform efficiency. They replace optimization with authorship. They give users structure that adjusts to them rather than the reverse. The demand for personalization has overtaken the demand for productivity. Software is shifting from uniform products to adaptive systems. People are choosing tools that fit them instead of tools that fit a model. The cracks are already visible. You can sense warmth under the surface. Curiosity is returning to the medium that once defined it.”

HOW TO STARTUP

Product management in the age of AI. Sudheendra Chilappagari, Neeraj Agrawal, and Jenny Kang of Battery Ventures argued that the role of product management in the AI-native era is primarily about moving quickly. “For years, engineering bandwidth was the limiting factor in product velocity. With AI-assisted coding tools, that bottleneck has shifted to surface a new one: clarity. Engineers can now build and ship features faster than PMs can plan, and in some cases, faster than users can absorb. As one product leader noted, customers are now asking them to slow down feature launches, frustrated that the UI seems to change every week. The greatest risk for product teams today isn’t moving too slowly; it’s moving fast in the wrong direction.”

HOW TO VENTURE

AI is the key to the exit. Jason Lemkin wrote up a great analysis of what is will drive future exits for SaaS businesses. In short, AI is the answer. “We’re in a very weird moment. There’s more capital available to invest than ever before—$476 billion in PE dry powder globally plus massive VC funds. But nearly all of it is chasing AI or the absolute best companies. The middle-tier SaaS companies—good companies, profitable companies, companies with happy customers—are in no-man’s land. But here’s the contrarian take, and Orlando Bravo has it right: AI valuations are in a bubble, but AI’s impact on enterprise software is real and massive. The $1.5 trillion software economy is getting a genuine tailwind. The firms and founders who understand this—who can add AI capabilities while maintaining profitability, who recognize that technology adoption is evolutionary not revolutionary, who build for the long term—will win the next decade.”

Groundcover is recognized in the 2025 Gartner Cool Vendors.

Aquant
Senior Enterprise Account Executive (NYC, Hybrid)

Groundcover
Infra Engineer (Tel Aviv)

Reco
Head of Revenue Operations (US, Remote)