Structured AI, a platform that features quality control and quality assurance artificial intelligence agents developed for construction, has raised a $4.2 million seed round, bringing its total funding to $5 million.
The round was led by FCVC. Other investors include Y Combinator, 20VC, Cherry Ventures, Zero Prime Ventures, Transpose Platform and Sequoia Scout. Construction documents are still the single source of truth that contractors rely on for most projects, and Structured AI’s optical recognition models focus on this workflow and analyzing what has been completed in the field against the documents.
“We trained our own computer vision models,” says Raymond Zhao, co-founder and CEO of StructuredAI. “Over the last two years, the vision has gotten really good and unlocking the identification of document marks and cross-checking with field work.”
The structured AI optical recognition platform performs quality control checks on entire sets of drawings before a senior engineer reviews them. Engineer Syska Hennessy Group has been testing the platform and helping StructuredAI develop its platform.
“About four years ago, we invested in them as one of our seed companies, and we’re also partnering with them kind of with the goal of seeing innovation and change in the industry,” says Robert Ioanna, executive vice president of Syska Hennessy Group in New York. “By partnering with some of these startups and giving them access, we get a tool that we really like and can use, that’s tailor-made.”
Syska Hennessy engineers told Zhao and the Structured AI team early in their partnership that MEP coordination was a common pain point they faced. This feedback eventually became the OCR and AI-based platform that addresses all kinds of QA/QC issues in projects.
“It shows some promise, and we’re moving forward and we’ve started to solve some pretty complicated things, like toilet drain plugs in the bathroom, and a lot of these things aren’t just fixing a line there or a line here in the documents, it’s more like analyzing something and interpreting something. They go to the next step, where it’s actually going to be in our drawing software, so I would literally check it like it’s a drawing software, like literally. where they’re going to recommend that a change be made and, then in your native drawing click accept and you’re done,” says Ioanna. Revit is the authoring tool that Syska has used for most of its Structured AI projects.
Looking for quick answers on construction and engineering topics?
Try Ask ENR, our new intelligent AI search tool.
Ask ENR →
Zhao says that while Structured AI is focused on speeding up QA/QC processes, it doesn’t want to take decisions away from architects or engineers.
“[Computer vision] it’s gotten really good at reasoning and doing logical thinking, but it doesn’t have the context of how an architect would interpret it,” Zhao says, adding that changes will still need to be approved by a person through system-generated notifications.
