Acelab raises $5.3M to transform architecture with ML search

You’d be forgiven for thinking of architecture as a high-tech industry. Oh sure, there have been numerous advances to the age-old practice of designing buildings, especially since the advent of the personal computer, and longstanding computer-assisted design (CAD) software programs like Autodesk‘s Revit and Graphisoft’s ArchiCAD make it far easier for architects to draft building blueprints, designs, floorplans, etc., including 3D models that the user can “walk through.”

But one area of architecture has remained frustratingly old school: searching for the actual supplies needed, such as windows, doors, roofing, and insulation — as they are typically dispersed across the websites and paper media of many different manufacturers. Thankfully, that is changing fast, according to Vardhan Mehta, a Harvard Master’s-educated architect turned founder and CEO of the startup Acelab, which is today announcing a new $5.3 million funding round to continue building out its intelligent building supplies search platform for his fellow architects.

“With building products being in the hundreds of thousands, it’s simply impossible for any architect or contractor to know about all of them,” Mehta wrote to VentureBeat in an email. “But every construction project is atypical in nature – due to code requirements, budget constraints, project location or client needs. Acelab machine learning (ML)-powered search engine is the brain of the platform. It gathers all relevant info from the architect, collates it with our proprietary database, and returns the strongest fits through an easy-to-use interface.”

The new funding round was led by Pillar VC, PJC and Draper Associates and joined by Alpaca, Steve Kaufer (Founder and former CEO of TripAdvisor), Erik Jarnryd (Former CEO of Harvey Building Products), Transcend Partners, Branagh Construction, Ken Lang, Klingenstein Fields, and Westview PE Fund.

Architects turned ‘hackers’: the origins of Acelab

“When I started working in architecture in 2017, I was surprised at how professional architects were still having to look through hundreds of manufacturers’ websites and brochures, and manually adding data on products to spreadsheets — they were kind of hacking their way through the problem,” Vardhan told VentureBeat in an exclusive video call interview.

Frustrated by this reality, Mehta and some of his fellow Harvard Master’s architecture students founded Acelab 2019 to provide a better solution. Together, they manually researched more than 39,000 products from hundreds of manufacturer websites and catalogs and combined them into a structured database. They further added 100,000 spec sheets from past architecture projects. Then, they worked with software engineers to develop their own ML search and recommendation engine, and provided a system for users to save, organize, and share their building products search results.

The result is Acelab’s three-tiered platform, which is currently used by more than 8,000 architecture firms, and offers the following features:

• ProductAdvisor, a proprietary, ML algorithmic search engine that includes building products from every major manufacturer in the U.S. and includes searchable attributes such as relative cost, sustainability, aesthetics, and more
• Project Workspace, a place where the architects and architecture firms who use Acelab Project can create lists of products for each project they are working on it and share it with colleagues, clients, building contractors, supplies manufacturers and other collaborators.
• a Collaboration Portal that connects a full-time Acelab product expert or a manufacturer’s specialist with the architects to provide more in-depth information on product availability and costs. Users can “instantly request quotes from suppliers, and share recommendations among multiple stakeholders,” Mehta said.

But Acelab goes further than just functioning as a “Google for architects.” Acelab’s ProductAdvisor generates product recommendations based on its learnings from other users on the platform — highlighting those products that they most often selected and added to their Workspaces, depending on their architecture project requirements.

“We often play the role of influencer,” Mehta said. “We guide the customer to which products and supplies to use.”

While users can’t order products directly from the platform itself (yet), they are only a click away from the manufacturers’ websites.

Acelab also employs a “data team” that manually canvasses building manufacturer websites and updates the product information on Acelab accordingly, though Mehta said the company has begun to automate some of this functionality.

“Typically, manufacturers revise their product lines every year or two,” Mehta said. “We do a manual refresh every 4-6 months.”

Mehta said that while the platform was built to reduce hours and hours of research time for architects, allowing them to focus on actually designing and communicating with their clients and manufacturers, the manufacturers he had spoken to were similarly enthralled by it.

“In the last four years, I’ve probably interviewed 450 manufacturers,” Mehta said. “A big challenge has been, for them, they traditionally relied on trade shows and catalogs to get the word out about their products, which didn’t always result in a successful connection if the architect didn’t need them at that moment. But with Acelab, we are able to make the connection between the manufacturers and the architects exactly when the architects are seeking them out.”

Acelab supports both residential construction, namely, homes and multifamily dwellings, and commercial construction products (offices, retail, hotels and convention centers).

Acelab’s blueprint for continued success

Newly flush with investment, Acelab has big plans to continue evolving its platform and add more helpful features for architects.

One big ticket item on the product roadmap: procurement, allowing users to buy products directly, which Mehta says is coming in the next 12-18 months.

The company further intends to add “historic price codes of manufacturers as reported by users” to its product search results, according to Mehta, allowing users to see more precisely the cost of each item, rather than the current relative pricing indicator (between one and four dollar signs, signifying increasing cost).

Acelab is also looking at adding a new generative AI feature that would automatically create specification documents for an architecture project with all the relevant product information already included, so an architect wouldn’t have to add them all one by one. Once an architect reviews the gen AI-created spec sheets for accuracy and additional information, they could then turn it over to the building contractors and client.

“Architects are facing enormous pressure to find the right products due to higher interest rates, supply chain delays, and labor shortages,” Mehta concluded. “Acelab helps the entire construction ecosystem find the products they need, so they can deliver projects on budget and on time.”