01 Aug 2026 4 min read Western Hemisphere

Dili Raises $21.7M From Khosla Ventures to Automate Infrastructure Compliance

Dili Founders Brian Fernandez and Anand Chaturvedi

Dili, a New York startup that automates regulatory compliance for federally funded construction and infrastructure projects, announced on July 30 that it raised a US$15 million Series A led by Khosla Ventures, bringing its total funding to US$21.7 million including a prior US$6.7 million seed round. Additional investors include Allianz, Rebel Fund, Brick and Mortar Ventures founder Darren Bechtel, and Y Combinator CEO Garry Tan. Founded in 2023 by CEO Anand Chaturvedi and CTO Brian Fernandez, both former Coinbase engineers, the company came out of Y Combinator's Summer 2023 batch. Chaturvedi previously built products at Coinbase generating more than US$50 million in revenue and conducted machine learning research at Apple.

Dili's platform uses large language models to extract structured data from unstructured project documents, then applies rule-based logic to check compliance against overlapping federal requirements including Davis-Bacon prevailing wage rules, Inflation Reduction Act apprenticeship provisions, OSHA safety standards, and EPA regulations. The company says it has processed more than US$1.4 billion in gross wages and roughly 16,000 certified payroll reports across over 700 federal projects, tracking 5.2 million labor hours, and has helped clients avoid more than US$50 million in fines and clawbacks. Customers include EDF, Radiance, Heelstone, and Borea. "A task that used to take a full day's work can now be dispatched in a matter of minutes," Chaturvedi said. The funding will go toward engineering, product, and go-to-market hiring, and toward broadening the platform beyond prevailing wage into wider audit and waste detection.

Market Context

Dili's pitch rests on a structural gap in how compliance is currently verified. Traditional auditors sample roughly 10% of project data on a semiannual basis, while Dili analyzes all of it continuously, cutting weekly review from more than seven hours to under five minutes by the company's account. Roughly half of Dili's clients license the platform as internal software, while the other half outsource compliance to Dili entirely, a dual model Chaturvedi expects to shift toward software over time as "software and AI are going to start eating a lot of those professional services workflows."

The demand backdrop is uneven. Data center construction has surged, with spending up nearly 139% year over year through late 2025 and average project costs approaching US$600 million, but data centers represent only about 6% of private construction spending and have not offset declines elsewhere. Manufacturing construction has contracted sharply as major semiconductor and battery plant projects move from heavy construction into equipment installation, and private nonresidential construction has declined for several consecutive months. That makes the composition of America's "infrastructure boom" narrower than the phrase suggests, concentrated in a handful of sectors and states.

The Signal

"A task that used to take a full day's work can now be dispatched in a matter of minutes." — Anand Chaturvedi, co-founder and CEO, Dili

Regional Relevance

For the United States: Prevailing wage compliance sits at the intersection of federal spending, labor protections, and construction economics, and Dili is building infrastructure that determines whether workers on publicly funded projects are actually paid the rates the law requires. Operating from New York, the company serves projects across energy, data centers, and advanced manufacturing, the segments absorbing the largest share of federal and tax-credit-supported investment. The US$50 million in fines and clawbacks the company says it has helped clients avoid points to how much compliance risk sits unmanaged in an industry where documentation errors, rather than deliberate underpayment, often drive penalties.

For the states absorbing the buildout: The construction activity Dili serves is geographically concentrated, with just five states, Louisiana, Texas, Virginia, Mississippi, and Pennsylvania, accounting for roughly 74% of year-to-date data center construction spending. In those states, contractors face compliance obligations on projects far larger than what local firms have historically handled, making automated audit capacity a practical constraint on which companies can bid competitively for federal and tax-credit-backed work. Weak conditions in the broader nonresidential market mean these projects carry outsized weight for regional construction employment.

The Other Side

What happens to a compliance business if the rules it automates get rolled back? On June 25, 2026, a federal judge in the Northern District of Texas vacated three provisions of the 2023 Davis-Bacon expansion after the Department of Labor declined to defend them, narrowing coverage for offsite manufacturing, delivery drivers, and retroactive contract application. The Davis-Bacon Act itself remains in force, and core obligations for certified payroll, worker classification, and apprentice ratios are unchanged, but industry groups including Associated Builders and Contractors have said "there is much more to be done" and are pursuing further litigation, leaving the regulatory surface Dili sells against genuinely in motion.

Is the "infrastructure boom" broad enough to support the growth this valuation implies? Manufacturing construction is contracting as semiconductor and battery projects wind down, private nonresidential construction has fallen for months, and the growth is concentrated in data centers, which account for roughly 6% of private construction spending; a downturn or pause in AI-driven data center buildout would narrow Dili's addressable market considerably.

Does a hybrid software-and-services model scale like software? About half of Dili's revenue comes from performing compliance work as a contractor rather than licensing a product, which delivers revenue now but carries service-business economics; the transition Chaturvedi describes, from outsourced work toward self-serve software, is the same shift many vertical AI startups have promised and comparatively few have completed at margin.