93 clients in 2 months. Now our estimating system is the bottleneck. 🚧🤖 We launched a construction estimating business in the Czech Republic roughly 2 months ago. So far: → 93 paying clients → Almost entirely construction companies → And we’re still just getting started. We basically act as an outsourced estimating department for contractors. A client sends us drawings, BoQs, specifications, PDFs, Excel files, photos, etc. We then: 1. Analyze the entire project 2. Determine exactly what needs to be priced 3. Extract quantities, materials and technical requirements 4. Find missing information / conflicts between documents 5. Build the bill of quantities / estimate 6. Price it using Czech construction cost databases + real market pricing 7. Finalize it in professional estimating software 8. QA the estimate and send it back to the contractor 9. We already use Claude, AI agents and our own custom-built internal app to help with this. And it works. But now we’ve reached the point where sales isn't our biggest bottleneck anymore — the estimating system is. We want to take what we have and build a MUCH better AI-assisted estimating workflow. Not a ChatGPT wrapper. Something that can reliably help us process real construction projects every single day — understand hundreds of pages of documentation, extract information, catch conflicts, prepare quantities, assist with pricing and eventually automate a large part of the estimating workflow while keeping human verification where it matters. I’m looking for someone in this community who REALLY understands one or more of these: - AI + construction estimating - Claude / LLM agent workflows - document & drawing analysis - quantity takeoff automation - estimating software - building reliable AI systems for real-world workflows Ideally, I’d love to find someone who could first look at what we've already built, tear it apart, tell us what we're doing wrong, and then potentially help us build the next version.