Why we built
Carrara Works.

Carrara Works exists because energy, development, and manufacturing have been underserved by the software industry. The people running these industries build housing, power the energy transition, and manufacture what we used to import. They have clear ideas about what their companies and industries should be doing differently. Instead of acting on those ideas, they are trapped in legacy software, regulatory paperwork, and administrative bloat. They spend their days reading contracts, evaluating tradeoffs, reconciling data across systems, and making decisions no piece of software they own can help them with. That's the gap Carrara Works exists to close: to unblock the people who already see what should change, and give them the software to ship it.

These industries have software for the data of record: vertical SaaS for solar design, project management for construction, ERPs across the board. None of it does the work above well. The judgment-heavy, context-bound work the operators spend their days on has stayed manual.

The reason is structural. The markets are too specialized: a few hundred or a few thousand companies share each workflow, not the millions horizontal vendors need. The workflows don't always repeat cleanly: the way one developer structures a tax credit transferability package is not the way the next one does, the way one interconnection engineer cross-references a FERC docket against an ISO queue position is not the way the next one does. These differences reflect different counterparties, financing structures, regulatory exposure, and decades of institutional preference. A vendor specialized enough to serve one of these workflows would lose the economics that made the model work in the first place. And until recently, no software could have done the judgment work.

What's changed is that AI, armed with the right tools and context, can now do work that used to require human judgment. Building software around this kind of work has gotten cheap: a small team in a quarter, where it used to be a product team in a year. The firms that capitalize will not be horizontal SaaS vendors. They will be generalists who understand the industry, embed with one client, learn how the work gets done, and build directly alongside the operators who do it.

Not every workflow is worth building this way. Software like this can only do the work when three conditions are met:

• Decision-heavy. The bottleneck is judgment, not data movement.
• Frequent. It runs often enough to justify the investment and accumulate signal.
• Context-bound. Today's models can do it, but only if the right documents, precedent, structured data, and tools are staged correctly.

Tax credit transferability sits cleanly in the center of all three. So does interconnection queue management. So do environmental permitting, regulatory filings, BOM cost engineering. Wherever decision-heavy, frequent, context-bound work is being done by hand, there is a workflow worth building.

We built Carrara, the workbench our team uses to develop, ship, and operate workflows for clients, so that the work of our generalists compounds. Carrara doesn't replace the software our clients already run. It sits on top, reading from their systems of record, running the judgment work, asking for human approval, writing back.

It handles tenant isolation, observability for production runs, and the runtime that keeps shipped workflows running long after an engagement. A generalist on their second engagement isn't rebuilding any of it. They spend their time on the only part that matters: understanding the client and the workflow. The product is a working, production workflow tailored to one client. The platform is what lets it get built fast enough and run reliably enough to be worth doing.

This isn't a new pattern for me. Before Carrara, I spent a year inside Greenlite, a permitting services firm, working across product, ops, and sales as revenue scaled 2.5x. At Bain, I spent ten months inside a $3B utility, streamlining its renewables acquisition (5GW pipeline, 750MW under LOI). Most recently at Moab, a vertical SaaS company, I've sat with CFOs and operations leaders at enterprise customers, shipped product into their teams, and built the production AI agents running on top. Carrara is what I would have wanted to build on the whole time.

If you run a function in energy, development, or manufacturing where the people who matter most to your business spend half their time on work that software should have eliminated by now, we should talk. We won't sell you a platform. We'll sit with your team, identify the workflows worth investing in first, and ship something useful inside months.

Every workflow we ship for one client lives in Carrara permanently. Every pattern we learn at one company sharpens what we ship at the next. Every operator we sit with is one more person freed from work that better software should have absorbed years ago. That's the work.

---

Hiring

We're looking for generalists with strong taste, comfort in ambiguity, and a real interest in the industries that build the physical economy. You don't need a background in energy or manufacturing. You need to want to learn what an interconnection engineer or a project finance lead does on a Tuesday, and to enjoy the kind of work that turns that understanding into software. If that's you, ben@carraraworks.com.