Analysis
CivilGrid has raised a $26 million Series A led by Spark Capital, with participation from Afore, A*, Ford Street Ventures, SNR and Energy Impact Partners, TechCrunch reported. The company was founded in 2020 by Josh Mackanic, who spent roughly ten years as an engineer at Pacific Gas and Electric -- the utility that also became an early proof point.
The product aggregates three data sets that have never sat in one place: underground utility records, property ownership, and environmental and permitting regulation. Engineering firms, municipalities and utilities use it to see what is beneath a street before they open it. The commercial hook is utility strikes and rework, which are among the most expensive line items in civil construction. In a PG&E case study, the platform surfaced more than $60 million in avoidable paving costs across 1,600 planned gas distribution projects.
Why now
Two forces are converging. The Infrastructure Investment and Jobs Act and the data-center buildout have both pushed enormous volumes of trenching, conduit and interconnection work into the same congested rights-of-way. At the same time, the underlying records are still overwhelmingly PDFs, paper as-builts and county GIS layers of wildly varying quality. That is a data-normalization problem, which is the kind of problem that has become tractable in the last three years and was not before.
The competitive landscape
The incumbent workflow is 811 "call before you dig" notification plus manual locating -- a regulatory process, not a planning tool. Adjacent venture-backed players include Exodigo, which raised $105 million in 2024 for subsurface imaging using sensors and AI, and Urbint, which does risk analytics for utilities. Bentley Systems and Trimble own the design-software layer. CivilGrid's position is upstream of all of them: not detecting what is underground with hardware, but assembling what is already known into something usable at the planning stage. That is a cheaper, faster-to-scale wedge, and a more defensible one if the data-integration work compounds.
The read for founders and GPs
Spark Capital leading a $26 million Series A in infrastructure data is a useful marker of where generalist capital is going as pure-play AI application rounds get crowded. The pitch here is not a model, it is a proprietary aggregation of public-but-unusable records, sold to buyers -- utilities and municipalities -- with slow procurement and near-zero churn once embedded. That combination has historically produced boring, durable, high-multiple businesses.
The obvious drag is sales cycle. Utility and municipal procurement measured in quarters is the reason most infrastructure-software companies grow linearly for four years and then compound. Six-year-old CivilGrid raising its A in 2026 is evidence of exactly that shape.
The regulatory tailwind nobody mentions
Utility strikes are not just expensive, they are increasingly a liability event. State public utility commissions have been raising penalties for damage to buried infrastructure, and after a decade of high-profile gas incidents, the documentation burden on excavation planning has grown steadily. A platform that produces an auditable record of what was known before a dig has value beyond cost avoidance -- it is evidence. That is the quiet reason infrastructure-data companies retain so well once installed: the customer cannot rip out the system without also losing the paper trail that protects them. It is also why the sales motion is slow. Someone in legal has to sign off, and legal is never the first meeting.
