4M Analytics

**Product and the concrete problem it solves.** 4M Analytics attacks one of the most expensive and dangerous blind spots in physical construction: nobody has a reliable map of what is buried underground. Every road widening, pipeline, transmission line, rail project, or data-center site sits atop a tangle of unmapped water mains, gas lines, fiber, power cables, and legacy utilities whose records are scattered across hundreds of municipal archives, utility owners, and as-built drawings -- if they exist at all. When a crew hits an unmapped line, the result is service outages, injuries, fatalities, litigation, and multi-week schedule slips; industry estimates put the cost of subsurface uncertainty in the tens of millions of dollars per day across US infrastructure. The legacy answer -- physically sending survey and "potholing" crews to locate utilities one dig at a time (the classic Subsurface Utility Engineering / 811 "call before you dig" workflow) -- is slow, partial, hazardous, and only covers the small footprint someone thought to check. 4M's answer is a software-first **Utility AI Platform** (marketed around its "4Map" product) that assembles a continuous, queryable, geospatially accurate map of the subsurface for an entire corridor or region *before* a shovel touches the ground, letting engineers see conflicts, quantify risk, and de-risk design during planning rather than discovering surprises during construction.

**Core technology and how it actually works.** The hard problem 4M solves is not a single sensor reading but data fusion at national scale. The platform ingests enormous volumes of heterogeneous, messy public and private evidence -- utility owner records, permits, as-built drawings, GIS layers, and high-resolution satellite and aerial imagery -- and reconciles them into a single coherent map. Computer-vision models detect above-ground utility "tells" (poles, manholes, valve covers, pedestals, meters, cabinets) from imagery; machine-learning and increasingly generative-AI pipelines then correlate those surface signatures with records and known network topology to infer the likely location, type, and confidence of buried assets. The output is not a raw document dump but a structured, risk-scored, CAD/GIS-ready deliverable that plugs into the engineer's existing design tools. In effect 4M is doing large-scale geospatial intelligence -- multi-source evidence fusion, feature extraction from imagery, and probabilistic reasoning about a hidden network -- which is precisely the discipline its technical founders practiced in uniform. The company frames coverage as nationwide across the US, a scale that manual field locating fundamentally cannot match, and its differentiator versus physical-scanning approaches is that software can map long linear corridors and whole regions remotely rather than one excavated pit at a time.

**Market, customers, and go-to-market.** 4M sells into the US infrastructure, engineering, and construction (AEC) market: state departments of transportation (DOTs), large engineering and design firms, utilities, and heavy-civil contractors. Its go-to-market leans on named, credibility-establishing customers -- public reporting and the company's own materials cite work with departments of transportation including **Georgia DOT (GDOT)**, plus major engineering and construction firms such as Stantec, HNTB, Halff, Westwood, MasTec, DPR Construction, Woolpert, and WSB. The value proposition to a DOT or design lead is concrete: fewer utility strikes, fewer change orders, faster permitting, and safer, more predictable schedules on projects where a single conflict can cost weeks. Reported operating scale is meaningful for a company of its size: 4M has said it supported more than 1,600 projects in a single year, mapping on the order of 400,000 miles of utilities across roughly 1.3 million acres. The strategic logic of relocating its headquarters to Austin, Texas in 2022 while keeping R&D in Tel Aviv is straightforward -- the buyer, the budget, and the regulatory context (US federal and state infrastructure spending) are American, so the company put its commercial center of gravity next to its customers.

**Traction, funding, and third-party validation.** 4M Analytics was founded in 2019 and is backed by a credible investor syndicate. Public reporting documents an $11 million early/Series A round, followed by a $30 million Series A extension in 2022 that brought the Series A to roughly $45 million and total disclosed funding to about $48 million. The extension was led by **Insight Partners** and **ITI Venture Capital Partners**, with participation from **Viola Ventures** and **F2 Venture Capital**. Notably, the cap table includes marquee Israeli mobility-mapping operators as angels -- **Noam Bardin**, former CEO of Waze (acquired by Google), and **Nir Erez**, founder and CEO of Moovit (acquired by Intel/Mobileye) -- a strong signal given that both built exactly the kind of large-scale, real-world mapping products 4M is attempting underground. The company reports roughly 100 employees. A calibrated reading: the traction (DOT customers, project volume) and investor quality are real and verifiable, but the most recent disclosed equity round is the 2022 Series A extension, so the current financing position, growth rate, and revenue are not publicly documented and are a central diligence question.

**Founders and team background.** The founding team is the company's sharpest asset and the root of its dual-use character. CEO and co-founder **Itzik Malka** served for years as an officer in the IDF's combat engineering corps -- the discipline whose job is literally to understand, breach, and build in complex terrain. CTO and co-founder **Yoav Cohen** is a former head of **Unit 9900**, the Israeli military's elite geospatial-intelligence unit responsible for satellite/aerial imagery exploitation, mapping, and visual intelligence -- one of the most sophisticated geospatial organizations in the world. COO and co-founder **Nir Cohen** brings combat-engineering and entrepreneurial experience. That pairing -- combat engineering (what is under and around the ground, operationally) plus 9900-grade geospatial intelligence (how to turn imagery and disparate data into a map) -- is an unusually precise fit for the problem of remotely mapping the subsurface, and it is what lends credibility to the claim that this is a defensible technology company rather than a records-aggregation service.

**Competitive dynamics.** 4M sits between two competitive fronts. (1) Against the entrenched manual status quo -- the traditional SUE / 811 field-locating and potholing workflow -- its edge is speed, corridor-scale coverage, and de-risking during design rather than during digging; the countervailing reality is that physical locating is the legally recognized, insurance-backed standard, and software inference must earn trust as a complement to, not yet a replacement for, boots-on-the-ground verification. (2) Against technology-first peers, the most direct comparison is fellow Israeli, IDF-alumni-founded **Exodigo**, which attacks the same "map the underground" mission but with a physical multi-sensor scanning platform (fusing ground-penetrating and other geophysical sensors with AI) and has raised dramatically more capital (well over $200 million). 4M's approach is asset-light and software-first -- mapping from existing records and imagery rather than deploying sensing hardware on site -- which is cheaper and more scalable across wide areas but is inference-based rather than direct measurement. The strategic question is whether the market rewards broad, fast, software-derived maps (4M) or high-fidelity, sensor-measured scans (Exodigo), or whether the two converge; 4M's capital disadvantage against a better-funded, adjacent competitor is a real risk.

**Defense, security, and resilience dual-use relevance.** 4M's dual-use case is a credible adjacency rooted in founder DNA and the fungibility of subsurface geospatial intelligence -- not a fielded defense product, and it should be read with that calibration. The fielded platform today is unambiguously commercial: civil engineering, DOTs, and construction. But the underlying capability -- turning multi-source imagery and records into an accurate map of hidden below-ground infrastructure -- is directly relevant to military engineering and homeland resilience: route and terrain analysis for maneuver, force-protection and base/FOB construction, protection of buried critical infrastructure (pipelines, power, water, telecom) that adversaries target, disaster-response and reconstruction after strikes or natural events, and the broader counter-tunnel/subsurface-awareness problem set that Israel in particular treats as a strategic priority. The founders' Unit 9900 and combat-engineering heritage means the team already thinks in these terms, and its Israeli peer Exodigo is explicitly discussed for tunnel-detection and defense adjacency. The honest framing: subsurface geospatial intelligence is inherently dual-capable, and 4M holds latent defense/resilience relevance, but the company has publicly productized only the commercial mission.

**Growth stage, trajectory, and key diligence risks.** 4M is a mid-stage company: founded 2019, roughly $48 million raised, about 100 employees, real DOT and enterprise customers, and demonstrated project volume -- clearly past seed/early, but not yet a mature, late-stage or public business. Its trajectory bet is that AI-driven, software-first subsurface mapping becomes a standard planning-phase layer for US infrastructure, riding sustained federal/state infrastructure spending. Key diligence risks: (1) **capital risk** -- no disclosed equity round since the 2022 Series A extension, so runway, revenue, and growth are undocumented and a well-funded competitor (Exodigo) is spending far more; (2) **accuracy/liability risk** -- inference from records and imagery carries confidence limits, and a miss on a live utility carries safety and legal exposure, so the product must be positioned carefully relative to legally-required physical locating; (3) **competitive/positioning risk** -- software-inference vs. sensor-measurement is an unresolved market question; (4) **concentration/procurement risk** -- DOT and large-firm sales are slow, reference-driven, and budget-cyclical; (5) **dual-use risk** -- the defense/resilience relevance is adjacency and founder-signal, not a demonstrated fielded capability, so any strategic-value case must not overstate it. The bull case is a capital-efficient, founder-strong geospatial-AI company owning a real infrastructure pain point with latent strategic relevance; the bear case is a records-and-imagery aggregator squeezed between the trusted manual standard and a better-capitalized sensing rival.