March 26, 2026
If you know RhinoDox, you know us as a bid management platform. What you might not know is what we've been building behind the scenes.
No press releases. No "we're disrupting construction" announcements. We spend our days in the trenches with subcontractors — listening to estimators, sitting in precon meetings, and building something we think matters.
We were too busy building it to talk about it. That changes now.
RhinoDox started as a bid management platform for specialty subs. We've processed over $20 billion in bids since 2020. We know what a Tuesday looks like for an estimator juggling five bids, three addenda, and a scope review that was due yesterday.
But the more time we spent with subs, the more we heard the same thing: "I know what to look for. I just don't have time to look for all of it."
That stuck with us.
So we started asking a different question. Not "how do we help subs read documents faster?" — every AI company in construction is trying to do that. Instead: "What if the tool understood what mattered at each phase of the bid, for each person who touches it?"
An estimator in bid development doesn't need the same information as a CFO reviewing contract terms. A precon director making a go/no-go decision doesn't need a 400-page analysis — they need five deal killers. And nobody needs to be told about risks that are industry-standard for their trade and their company.
That insight became the foundation for LARRY.
LARRY is a preconstruction intelligence platform. Not a document reader — there are enough of those. LARRY reads everything in your bid package — specs, contracts, safety manuals, OCIP documents, insurance requirements, environmental compliance — and tells you what actually matters for this bid.
Here's what makes it different:
It reads across document types, not just specs and contracts. Most tools stop at the spec book. But a $500K miss caused by conflicting requirements between a project spec and a manufacturer installation guide doesn't care that they're in different documents. Nobody catches cross-document conflicts because nobody reads all the documents together. That's the problem we set out to solve.
It filters by where you are in the bid. At go/no-go, it shows deal killers. During bid development, it pulls scope gaps, cost drivers, and conflicts. Pre-award, it flags the contract terms that'll bite you. Same project, different lens depending on when you're looking.
It cites everything. Every finding links to the source document and page number. We interviewed twelve subs and every single one said the same thing: "Don't tell me there's a risk — show me where it is in the document." Trust and verify. That's how construction works.
Building LARRY wasn't just an engineering project. It was an education.
A 38-year construction industry veteran told us the "grand slam" is sorting outputs by estimating, legal, risk matrix, and CFO. Different people need different signals — showing everything to everyone is noise.
We listened to a drywall estimator who runs five bids simultaneously describe his ideal workflow: "Dump the docs Tuesday morning. Come back in an hour. See what matters." That became our benchmark for speed.
We watched a project executive describe — unprompted, three sessions in a row — the moment after awarding a job when he wants to cross-reference the contract against what he bid. "Why wouldn't you prompt me to do that?" That conversation told us exactly what to build next.
We heard a two-estimator mechanical shop explain that for them, go/no-go isn't an optional summary — it's survival. "We pick and choose what jobs we want to take on. How much burden do I want?" When you can only bid ten projects a month, saying no to the right ones is the highest-leverage decision you make.
And we learned that five out of twelve subs are already uploading specs to ChatGPT to try to solve this. The demand is there. The behavior exists. What's missing is a tool built specifically for preconstruction — with citations, with cross-document intelligence, and without hallucinations.
We could have kept building for another year. There's always more to add. But the subs we're working with in pilot — mechanical, electrical, drywall, plumbing — are telling us the same thing: this is already changing how they bid.
And the industry isn't waiting. Every sub in the country is going to be using AI for preconstruction within two years. The question isn't whether — it's whether the tool they use actually understands their business, or whether it's another generic platform that dumps everything and lets them sort through it.
We built LARRY because we believe AI is not the answer. Understanding the problem is the answer. Then you apply AI.
We've been building. Now we're ready to show you what we built.
If you're a specialty sub and you want to see what we've built — reach out. justin.ullman@rhinodox.com or Colin French at colin.french@rhinodox.com.
Justin Ullman is the Founder & CEO of RhinoDox. $20 billion in bids processed. In the trenches with subcontractors since 2020.
