As new data centers and power infrastructure come online, the compliance burden around construction is becoming part of the infrastructure story. Dili, a new AI compliance company, is positioning itself around that problem with fresh funding and a focus on US projects tied to complex rule sets.
A new round for infrastructure compliance
On Thursday, Dili announced $15 million in series A funding. The round follows a $6.7 million seed round and brings the company’s total fundraising to $21.7 million.
The Series A was led by Khosla Ventures. Allianz, Rebel Fund, Brick and Mortar Ventures’ Darren Bechtel, and Y Combinator’s Garry Tan also participated. Dili was previously part of Y Combinator’s Summer 2023 batch.
The company is not pitching AI in the abstract. Its target is the administrative and regulatory work around the new crop of US infrastructure projects, where construction documents, payroll systems, vendor records and other internal information can all matter for compliance.
Why construction rules are hard to manage
AI for compliance is already a familiar startup category, but Dili is focused on the particular complexity of construction projects. That complexity can increase when projects receive some kind of federal funding.
Dili co-founder and CEO Anand Chaturvedi pointed to Davis-Bacon rules as one example. Those rules allow the Department of Labor to set prevailing wages for certain projects. A separate set of prevailing wage and apprenticeship rules, or PWA rules, apply to clean energy projects funded under the IRA.
Other rules can also enter the picture depending on the work involved. The source describes OSHA or EPA rules as applying in various cases based on the nature of the project.
The practical challenge is that these requirements can overlap. A project may need to satisfy wage, apprenticeship, workplace, environmental and funding-related obligations at the same time. For companies managing large builds, the issue is not only knowing the rules, but checking the right project information against those rules before a small error becomes expensive.
Chaturvedi said non-compliance can lead to millions of dollars in fines for those projects. That is why Dili’s pitch emphasizes checking all incoming information rather than relying only on sampled data.
How Dili uses AI without leaving the final call to AI
Reliability is central to the product because compliance work cannot depend on loose interpretation. According to Chaturvedi, Dili’s architecture is designed so that contemporary AI models are used in the company’s data layer, not as the final compliance decision-maker.
That data layer takes unstructured documents and turns them into structured data. Once the information is structured, a deterministic system sorts it according to compliance rules that are complex but static.
This distinction matters. The AI handles the messy work of reading across documents and systems. The rule application is then handled by a more fixed process, which is intended to reduce the fuzziness that can come with large language models.
In the company’s framing, the system can read across internal documents, vendors’ documents, ERP information and payroll systems information. From that broader context, it pulls out the data needed for reporting or compliance.
When the system works as intended, a task that once took a full day’s work can be completed in a matter of minutes. The source does not describe that as replacing the importance of compliance judgment; rather, the value is in making the information review faster and more complete.
Where the software is being used
Dili says its software is already being used at about 700 projects. Those projects range from manufacturing facilities to data centers.
The customer model is split in two. Chaturvedi said roughly half the projects use Dili as an in-house software tool. The other half outsource the entire compliance process on a contractor model.
Dili can support both arrangements. Still, Chaturvedi expects the industry to move more toward the software model in the years ahead. His view is that software and AI will absorb more professional services workflows, causing more customers to bring those workflows in house.
That shift, if it happens, would make compliance software part of the operating layer for infrastructure development. Instead of treating compliance as a periodic outside review, companies could use software to continuously organize and check the information that already flows through their projects.
What this signals for the infrastructure boom
The growth of AI is driving demand for data centers and power infrastructure. Dili’s funding suggests a related market is forming around the paperwork, reporting and rules that come with building that infrastructure.
The central idea is straightforward: more projects mean more documents, more vendors, more payroll data and more compliance exposure. If those inputs remain fragmented, mistakes can be hard to catch early. If they can be structured and checked systematically, project teams may have a better chance of finding issues before they become costly.
Dili’s bet is that AI can help with the first part of that problem by extracting the relevant information from messy sources. Its deterministic system is meant to handle the second part by applying the compliance logic. For infrastructure projects with federal funding or clean energy requirements, that combination is the company’s case for bringing AI into a traditionally service-heavy workflow.