A startup called Dili has raised $15 million in a funding round led by Khosla Ventures, with additional investment from Allianz, Rebel Fund, Darren Bechtel of Brick and Mortar Ventures, and Garry Tan of Y Combinator. The round brings Dili's total funding to $21.7 million, including an earlier $6.7 million seed round. TechCrunch

Dili, founded by CEO Anand Chaturvedi and based at dili.com, makes software that helps construction companies manage the paperwork and rules that come with federally funded infrastructure projects. As of July 2026, the company's platform is in use across roughly 700 projects. TechCrunch

The main problem Dili tackles is something called Prevailing Wage and Apprenticeship, or PWA, compliance. When a construction project uses federal money, the government requires that workers be paid the going rate for their trade in that local area. Contractors also have to meet certain apprenticeship standards. Dili's software helps users follow these rules, which come from the Davis-Bacon Act and the Inflation Reduction Act, and it produces compliance reports that are ready to submit to the IRS. Dili PWA

Y Combinator's company directory describes Dili as an AI system of record for physical industries, meaning it serves as the central, authoritative place where a project's compliance data lives. The company's LinkedIn page calls the product an AI co-pilot that automates due diligence and deal management for tax credits. Y Combinator LinkedIn

Dili was part of Y Combinator's Summer 2023 batch, and YC's Garry Tan participated in the current funding round. TechCrunch Y Combinator

The lead investor brings a notable profile. Vinod Khosla, whose firm led the round, was ranked No. 10 on the Forbes 250 list of America's Greatest Innovators in 2026. Khosla Ventures

The mix of investors here is telling. Allianz is a major insurance company, and Brick and Mortar Ventures focuses on construction technology. Their participation alongside Khosla Ventures, which typically backs deep technology, reflects the fact that federal infrastructure compliance sits at the intersection of labor law, tax credits, and insurance risk. The investor list covers each of those areas rather than representing a single bet on AI as a general-purpose tool.

The rules involved are genuinely complicated. The Davis-Bacon Act requires contractors on federally funded public works to pay locally prevailing wages, and what counts as prevailing varies by location and trade. The Inflation Reduction Act added further requirements on top of certain clean energy projects to qualify for enhanced tax credits. For contractors dealing with both sets of rules, the paperwork is substantial, and mistakes can lead to withheld payments or even disqualification from federal programs.

Dili's ability to generate IRS-ready compliance reports directly from project and payroll data is where AI and large language models can genuinely cut down on manual work. Think of it like having a system that reads your payroll records and project files, understands the relevant regulations, and produces the forms you need without someone typing them out by hand. The company's positioning as a system of record means it aims to be the trusted, central source of truth for a project's compliance status, rather than just a tool that generates individual documents.

The adoption numbers are worth noting. Seven hundred active projects as of July 2026 suggests the product has moved beyond early testing into real-world use, though the size and complexity of those projects is not detailed in the available reporting.

For construction contractors, the appeal of an automated compliance system is practical: it shortens the time between when work is done, when payroll is verified, and when the required paperwork reaches the government. That matters because manual compliance processes tie up money and time. Whether Dili becomes the go-to platform for this segment will depend on how well it handles the many local variations in wage rules, and on whether its records hold up when federal agencies audit or verify the data.