Pune: For 32-year-old Lesh Dixit, a delayed home renovation proved to be a boon. It provided him deeper insights into a larger problem which spurred him on to fix it.
“It works with the drawings, photos and messy inputs construction already produces. Once project activity, procurement, payments and vendor relationships are visible digitally, the contractor becomes easier to assess. Financing can then be embedded directly into the point at which the business requires, such as purchasing materials.
“AI does not require contractors to learn something and structure their data,” Dixit said.
“Every design decision we got right came from watching, not from a product meeting. That approach affects hiring too. “It requires sustained investment in data, model work and people who understand construction deeply enough to teach a system what a good estimate actually looks like. Traditional software often requires contractors to alter behaviour, enter structured data and learn new workflows. AI potentially reverses that equation by allowing software to work with information construction teams already produce — drawings, photographs and other relatively unstructured inputs.
Powerplay began with a small team and has since grown to over 100 employees.
“What it really took was proximity,” Dixit said. But, according to Dixit, construction technology requires more than conventional software talent. “Building domain-trained agents for estimation, procurement and risk is not a software problem you solve once,” he said.
