Archiboost AI Raises $6.2M for AI Review of Construction Documents Before They Reach the Field
The San Francisco startup flags errors, omissions and coordination issues across drawing sets and specs — the second construction-document-AI funding round within a week.
Archiboost AI, a San Francisco startup that exited stealth in March 2026, announced a combined pre-seed and seed round of US$6.2 million from Hanabi Capital and Marathon Venture Capital on November 3. The company's platform reads PDFs and Revit metadata to find discrepancies between drawings, specifications and code requirements before documents are issued for construction.
The company claims more than 550,000 drawing sheets reviewed across projects worth $3.5 billion, over 120,000 flagged issues, up to six times more issues detected than manual QA/QC, and 70 percent fewer field coordination problems — all vendor-reported figures without independent verification. Plans include a construction administration agent drawing data from Autodesk Construction Cloud and Procore; clients include WRNS Studio, OTJ Architects and RXR.
The round lands one day after Shackleton AI's $5 million raise, meaning two construction-document-AI startups were funded within four days. Venture capital is consistently betting that document validation before field execution becomes a standard layer of construction workflows — including in markets beyond the US.