ParspecSTART -UP using artificial intelligence to enhance the efficiency of the construction industry supply chain, collected $ 20 million in financing of the A series, said only Crunchbase News.
The company goals to assist sales agents and wholesale distributors “effectively bid and deliver” construction products. The company claims that its key distinguishing feature is the possibility of “immediately identifying products available on the market, which meet the complex specifications provided by the customer” using artificial intelligence.
“This allows clients to reduce time and costs in half, while improving the quality and compliance of offers, enabling them to bid and win more projects”, CEO and co -founder Forest flags He said in an interview.
The company began concentration on lighting and electrical products and has since expanded to support mechanical, electrical and hydraulic products (MEP).
Thresholds (earlier DFJ) was funding for the PARSPEC AS series, with the participation of existing supporters, including Innovative projectsIN Construction undertakingsIN Heartland Ventures AND Hometeam Ventures. Financing is brought by San Mateo, in California, in California, it has increased to $ 31.5 million so far.
While the company refused to reveal its valuation, Flager told Crunchbase News that it has been “about 3x” since the time of seed increases in March 2022.
The increase appears at a time when the financing of undertakings for start-ups related to real estate in the financing of real estate in the United States has dropped. The sector recorded a total of 13.7 billion dollars as part of 1088 offers in 2021, for Crunchbase. Wounded by the increase in the rates of interest of the mortgage, real estate prices and construction costs, the financing of real estate fell to $ 4.9 billion in 2023, before it fell to $ 3.6 billion on 2024. In 2025, $ 2.1 billion entered the startups of real estate in 201 offers in the USA in the USA in US
Optimization through information
Flags with a co -founder Pratush Havela While working on a doctorate on Stanford University. Havela was a student in her research laboratory. By conducting research on how designers can use calculations to higher examine the options and optimization of buildings in terms of energy efficiency, the couple realized that “optimizations are as good as the information in which you provide,” said Flager.
“Information about building products, such as the efficiency of different windows or lighting materials, was not really available on a scale,” he added. “It was simply very difficult to get this information.”
Flager was led by a software team CherryAND Softbank-It’s building technological fact, which ultimately went through in 2021, when he obtained a higher understanding of mechanics, how the material is purchased, in addition to information that the distributor has.
“I wanted to build a solution that strengthened the distributor, and ultimately provide funds for collecting product information in a way that in my opinion can be useful for many different players in the value chain,” he recalls.
Parspec was turned on in 2020, but Flager and Havelia began to focus on it full -time in 2021.
Flager claims that today there are no other products on the market that are able to right away discover products in accordance with SPEC or mechanically locate product documentation, akin to installations or guarantees. He believes that this distinguishing feature, in combination with the launch of the citation product in January 2024, gives the company a competitive advantage.
As a result, Parspec has noted an increase in 4x revenues in the last 12 months.
Flager believes that Parspec – which mixes the words “parsia” and “specifications” – has the ability to cut back construction costs in the supply chain. The prices of building materials are about 40% higher than in 2020, so more agile and efficient construction supply chain is more needed than ever.
You have a component
Describing Parspec as a “workflow tool”, Flager said that every component in the flow of work uses AI’s capabilities. The company uses machine learning and LLM models to extract products from unstructured documents, akin to drawings and specifications. Then identifies the technical requirements from the natural language specification.
“Then, when we talk about matching products to the specification, we have the entire AI pipeline to create a comprehensive pipeline in which we crawl around 4,000 different web sites of manufacturers, indexing all PDF documentation that they have there, and then we use AI ML models to drag out attributes and create a catalog that matches these products.
Parspec currently has 288 customers, mainly wholesale distributors and sales agencies in the MEP products industry. The company said that it cooperates with 4 of the five largest electrical distributors and three of the five largest lighting agencies in the USA
His revenue model is a subscription based on use. The subscription use component is based on the number of documents that the customer creates using the Parspec platform. Qualified documents include quotes, applications in addition to operational and maintenance packages.
Ventures threshold partner Mo Islam Crunchbase News said through e -mail that his company was impressed by the “truly native AI AI product to automate the order process”, which in line with him allows customers of the company to “win more activities, while improving the profit margins in the billion billion industry.”
“A set of data on the materials of reserved Parapec materials and refined AI generative models give the company a strong moat of data and AI competitive advantage,” added Islam. “The company’s platform can automate the entire flow of work in the construction industry, from the choice of product and prices to quotation and transmission.”