Verrus

Hyperscale data centers with integrated battery energy storage and energy-management software for grid services.

Cover Block

Open sources

Field	Detail
Name	Verrus
Tagline	Hyperscale data centers with integrated battery energy storage and energy-management software for grid services.
Headquarters	Brooklyn, New York, United States [LinkedIn, retrieved 2026]
Founded	2023 [Tracxn, retrieved 2026]
Stage	Seed [Tracxn, retrieved 2026]
Business model	B2B
Industry	Cleantech / Climatetech
Geography	North America
Growth profile	Venture Scale
Founding team	Corporate Spinout [TechCrunch, March 2024]
Funding label	Undisclosed [Tracxn, retrieved 2026]
Total disclosed	~$801,000 [Tracxn, retrieved 2026]

What an Investor Needs First

PUBLIC Verrus is building hyperscale data centers that pair onsite battery energy storage with energy-management software, a configuration that merits attention because it targets the two constraints now shaping AI infrastructure buildouts: grid capacity and power flexibility [TechCrunch, March 2024] [Sustainable Tech Partner, March 2024]. The company emerged from stealth in March 2024 as a spinout from Sidewalk Infrastructure Partners, with public materials describing a model in which data centers can serve both compute demand and grid services, rather than treating energy infrastructure as a separate layer [TechCrunch, March 2024] [Verrus, retrieved 2024]. Its stated differentiation is an energy-aware architecture that aims to make facilities "grid aware, carbon aware, and compute aware," with public claims around flexible workload balancing, carbon-free energy procurement, and support for both conventional and AI-oriented workloads; those claims are directionally interesting, though much of the underlying performance evidence remains company-described or trade-press reported [Sustainable Tech Partner, March 2024] [Verrus, retrieved 2024] [OSTI.GOV, retrieved 2026].

The team reads as infrastructure-first rather than software-first: Nelson Abramson has been identified as CEO and previously led infrastructure at X, while SIP co-founders Brain Barlow and Jonathan Winer are publicly tied to Verrus leadership, and commercial leadership includes Jeff Monroe, previously associated with Verne Global [Sustainable Tech Partner, March 2024] [Global Corporate Venturing, April 2024] [TechCrunch, March 2024]. Title consistency is not fully resolved in the public record, with different sources describing Abramson, Barlow, and Winer in overlapping chief executive roles, which investors should clarify directly during diligence [neuron.com, retrieved 2026] [LinkedIn, retrieved 2026] [Global Corporate Venturing, April 2024]. On financing, public databases point to limited disclosed funding at the company level, while the more material backing appears to come through SIP and its investor base, including Alphabet, Ontario Teachers' Pension Plan, and StepStone Group; Tracxn lists a $119,000 round in June 2023 and an undisclosed financing in December 2024, but those entries are lightly corroborated [Verrus, retrieved 2024] [Tracxn, retrieved 2026].

The business model is B2B and capital intensive by design, with TechCrunch reporting that a single hyperscale facility could cost roughly $1 billion and require hundreds of millions in equity, which frames both the upside and the execution burden [TechCrunch, March 2024]. Over the next 12 to 18 months, the public watchpoints are straightforward: whether Verrus converts its NREL-linked demonstration work into confirmed commercial deployments, whether prospective sites in Arizona, California, and Massachusetts advance beyond planning, and whether the company can show named customers or contracted capacity rather than concept validation alone [Latitude Media, retrieved 2026] [OSTI.GOV, retrieved 2026] [TechCrunch, March 2024]. Partially corroborated -- The section relies on a mix of independent reporting, company materials, and one database entry for funding that is only partially corroborated.

Taxonomy Snapshot

Axis	Value
Stage	Seed
Business Model	B2B
Industry / Vertical	Cleantech / Climatetech
Geography	North America
Growth Profile	Venture Scale
Founding Team	Corporate Spinout
Funding	Undisclosed (total disclosed ~$801,000)

Inside the Company

PUBLIC

Verrus presents as a recent infrastructure spinout rather than a conventional software startup. The company is headquartered in Brooklyn, New York, and public materials describe it as developing hyperscale data centers that pair onsite battery storage with proprietary energy-management software [Verrus, retrieved 2024]. Its public positioning matters because it places the company at the intersection of data-center development, power infrastructure, and grid services, a capital-intensive lane that tends to move on long development cycles rather than quick product iteration [Verrus, retrieved 2024].

The public chronology is still thin, but a few milestones are clear. Verrus was founded in 2023, according to the structured company record, and by 2024 its website and hiring materials were already framing the business around "next generation, greenfield hyperscale data centers" and AI-linked workload flexibility [Verrus, retrieved 2024]. A June 30, 2023 financing entry of $119,000 and a December 2, 2024 financing entry with undisclosed amount appear in Tracxn, though neither filing is well described in public source material and should be treated cautiously [Tracxn, retrieved 2026].

The legal-entity picture is not fully resolved from the allowed source set, but the operating identity is visible. Verrus's website ties the company to Sidewalk Infrastructure Partners and frames the business as a developer and operator of integrated, battery-backed data-center infrastructure [Verrus, retrieved 2024]. On the available public record, the company appears to be early in commercialization, with the clearest external markers being its web presence, recruiting activity, and sparse funding traces rather than a long list of operating-site milestones [Verrus, retrieved 2024] [Tracxn, retrieved 2026].

Partially corroborated -- This section relies primarily on Verrus website materials, with partial corroboration from Tracxn on timing of funding entries.

Under the Hood

MIXED

Verrus is positioning the product at the intersection of data center design and power-market participation, which is more specific than a generic sustainability claim. Public materials describe "next generation, greenfield hyperscale data centers" built with onsite battery energy storage and proprietary energy-management software that coordinates power, carbon, and compute requirements [Verrus, retrieved 2024]. Trade coverage is directionally consistent: Sustainable Tech Partner reported that the system is meant to make facilities "grid aware, carbon aware, and compute aware," so workloads can be balanced against grid conditions and carbon-free energy availability [Sustainable Tech Partner, March 2024]. TechCrunch's March 2024 launch coverage framed the architecture around battery-backed microgrids and flexible operation for AI and machine learning workloads, especially those that can tolerate variable uptime [TechCrunch, March 2024].

The technical promise rests on controlled flexibility rather than raw compute novelty. Verrus says its electrical architecture can provide continuous power when required, while allowing more elastic power consumption for AI training and other batch workloads, and it claims up to 99.999% availability when customers need it [Verrus, retrieved 2024]. The same materials describe optional day-ahead workload shedding during periods of peak cooling strain, with the company claiming that this can open capacity that would otherwise be unavailable for 95% to 99% of the year [Verrus, retrieved 2024]. Those are still company-side performance claims, but they map to a clear product thesis: a hyperscale facility that behaves partly like critical infrastructure and partly like a dispatchable grid asset.

A verified public demo gives the concept more weight than website language alone, though it does not yet establish broad commercial deployment. OSTI.GOV documents Verrus validation work on Vulcan, a 70 MW utility-scale test platform powered by NREL's ARIES Virtual Emulation Environment [OSTI.GOV, retrieved 2026]. Latitude Media separately reported that Verrus used a software testing platform connected to 70 MW of infrastructure at NREL's Colorado labs to test performance under grid conditions, and datacenters.com said the company is moving from proof-of-concept toward commercial deployments [Latitude Media, retrieved 2026] [datacenters.com, retrieved 2026]. Two open roles also suggest the technical stack spans facility-level systems and software operations, with one Product Manager posting focused on data center controls and one Staff Site Reliability Engineer role in California, which supports a software layer view, though any stack detail beyond that remains inferred from job postings [Verrus, retrieved 2024] [LinkedIn, retrieved 2024].

Reasoned from indirect evidence -- Core product descriptions rely materially on company materials, with partial external corroboration from TechCrunch, Sustainable Tech Partner, OSTI.GOV, Latitude Media, and datacenters.com.

Market Research

PUBLIC The market matters now because AI-driven data-center demand is colliding with power-system constraints, and Verrus is positioned at that intersection rather than only inside the colocation market [TechCrunch, March 2024] [Latitude Media, retrieved 2026].

The available public record does not support a conventional TAM, SAM, and SOM build for Verrus from named third-party market reports, so the cleaner approach is to anchor on the cost and capacity problem the company is trying to address. TechCrunch reported in March 2024 that Jonathan Winer said a hyperscale Verrus facility could cost about $1 billion, with "hundreds of millions of dollars of equity" required, which places the company in a capital-intensive slice of digital infrastructure rather than software-scale energy management alone [TechCrunch, March 2024]. That matters for market definition: the relevant opportunity is the overlap between hyperscale data-center development, onsite storage-enabled power infrastructure, and grid-services participation, not a narrow demand-response software category [TechCrunch, March 2024] [Sustainable Tech Partner, March 2024].

Demand drivers in the cited research are straightforward. Verrus emerged publicly with a pitch tailored to AI and machine-learning workloads, especially workloads that can tolerate some flexibility in uptime or scheduling, while still supporting high-availability requirements when needed [TechCrunch, March 2024] [Verrus, retrieved 2024]. Sustainable Tech Partner described the system objective as making data centers "grid aware, carbon aware, and compute aware," a framing that ties three pressures together: rising compute demand, stricter expectations around carbon-free energy procurement, and the practical difficulty of securing enough power at new sites [Sustainable Tech Partner, March 2024]. Latitude Media's reporting on the NREL demonstration adds a second tailwind, namely growing utility and grid-operator interest in flexible large loads that can respond to real-world grid conditions rather than behave as fixed consumption [Latitude Media, retrieved 2026] [OSTI.GOV, retrieved 2026].

The nearest adjacent markets are utility-scale battery storage, conventional hyperscale and colocation development, and industrial demand flexibility. Verrus is not presented publicly as a standalone battery developer, and it is not framed as a pure colocation provider either; the proposition is a bundled infrastructure model where battery storage and software change the economics and operability of the data center itself [TechCrunch, March 2024] [Verrus, retrieved 2024]. Substitute approaches therefore include standard hyperscale builds paired with grid interconnection and backup systems, as well as separate procurement of utility demand-response or carbon-matching tools, but the company's argument is that these functions are more valuable when designed into the facility from the start [Sustainable Tech Partner, March 2024] [TechCrunch, March 2024].

Macro and regulatory forces cut both ways. On the supportive side, public concern about whether data-center growth can strain regional grids appears to be creating space for models that can modulate load, store energy onsite, and participate in power markets, which is consistent with Verrus's NREL validation effort on a 70 megawatt test platform [Latitude Media, retrieved 2026] [OSTI.GOV, retrieved 2026]. On the constraining side, the same sources imply a long development cycle, heavy upfront financing needs, and exposure to interconnection, permitting, and electricity-market rules that vary by region, especially if the first sites are spread across Arizona, California, and Massachusetts as discussed publicly [TechCrunch, March 2024].

Cited sizing or infrastructure claim	Figure	Context
Hyperscale Verrus facility cost	~$1 billion	Approximate project cost cited by Jonathan Winer [TechCrunch, March 2024]
NREL-linked test infrastructure	70 MW	Test platform used to validate grid-interactive behavior [Latitude Media, retrieved 2026] [OSTI.GOV, retrieved 2026]
Optional capacity availability claim	95-99% of the year	Company claim for customer access to otherwise unavailable capacity under workload-shedding design [Verrus, retrieved 2024]

The table shows why this is better read as an infrastructure-market thesis than a typical startup software wedge. The public numbers are sparse, but the ones that exist point to large asset value, meaningful power scale, and a market shaped as much by grid constraints as by compute demand.

Reasoned from indirect evidence -- This section relies on one independent news report for project economics, one independent report plus one technical report for the 70 MW validation context, and one company source for performance-related market claims [TechCrunch, March 2024] [Latitude Media, retrieved 2026] [OSTI.GOV, retrieved 2026] [Verrus, retrieved 2024].

Competition and Substitutes

MIXED Verrus is not competing head-on with standard colocation on price alone, at least in the public record. The company is positioning itself at the intersection of hyperscale data-center development, battery-backed power systems, and grid-responsive energy software, a lane that is closer to infrastructure design than to conventional hosting services [TechCrunch, March 2024] [Verrus, retrieved 2024].

The competitive map is therefore best read in three layers. Second are challengers building energy-aware compute infrastructure, where Verrus is trying to differentiate through integrated battery storage and software that makes facilities, in its words, more "grid aware, carbon aware, and compute aware" [Sustainable Tech Partner, March 2024]. Third are adjacent substitutes: enterprises can still meet AI and batch-compute demand through conventional hyperscale leases, cloud availability zones, or less integrated behind-the-meter energy upgrades, all of which reduce the need to adopt a new facility architecture before it is commercially proven [TechCrunch, March 2024].

The clearest edge visible today is sponsor quality and domain overlap, not operating scale. Verrus emerged from Sidewalk Infrastructure Partners, and public materials tie that parent platform to Alphabet, Ontario Teachers' Pension Plan, and StepStone Group, which matters because the company's own concept appears unusually capital intensive for a startup, with Jonathan Winer telling TechCrunch that a hyperscale facility could cost about $1 billion and require hundreds of millions of dollars of equity [Verrus, retrieved 2024] [TechCrunch, March 2024]. The team also appears built around infrastructure and power experience, with public sources linking Nelson Abramson to X infrastructure, Jeff Monroe to Verne Global, and Jeff Bladen to prior grid and power roles [Sustainable Tech Partner, March 2024] [Latitude Media, retrieved 2026] [metroconnectfall.com, retrieved 2026]. That edge is real but perishable. Capital access and operator credibility help open doors with utilities and large compute buyers, yet they are not the same as an installed base, signed anchor tenants, or a repeatable construction program.

The exposure is equally plain. Verrus has public evidence of a technology demonstration with NREL's Vulcan 70 MW test platform, which is useful technical validation, but the available sources do not identify a named paying customer, an operating commercial site, or a signed utility program [OSTI.GOV, retrieved 2026] [Latitude Media, retrieved 2026] [TechCrunch, March 2024]. That leaves the company vulnerable to any established operator that can add similar battery integration, demand flexibility, or procurement software to an existing campus and sell it through already owned enterprise channels. In other words, the main competitive threat is not a better publicized startup in this dataset, but an incumbent's ability to absorb the Verrus thesis into a broader delivery machine before Verrus converts concept into contracted capacity.

Over the next 18 months, the most plausible competitive scenario is a bifurcation between technical credibility and commercial proof. Verrus is the likely winner if grid constraints tighten further and AI workloads continue to reward flexible power architecture, because its public pitch is designed around exactly that bottleneck [TechCrunch, March 2024] [Sustainable Tech Partner, March 2024]. Verrus is also the likely loser if buyers prioritize delivered capacity over architecture novelty, since incumbents with existing campuses, utility relationships, and procurement teams can meet demand without asking customers to underwrite a new development model. The near-term contest is therefore less about who has the strongest idea and more about who can turn power-aware design into bankable, financeable projects first.

Partially corroborated -- This section relies on public sources from TechCrunch, Sustainable Tech Partner, Verrus materials, Latitude Media, and OSTI.GOV.

Opportunity

PUBLIC The prize here is unusually large: if Verrus can prove that battery-integrated, grid-responsive data centers deliver lower-cost AI capacity without sacrificing reliability, it could become a scarce developer-operator for a new class of power-constrained digital infrastructure [TechCrunch, March 2024] [Sustainable Tech Partner, March 2024].

The headline opportunity is not simply to build another data center company. It is to define a design standard for hyperscale facilities in regions where power access, carbon intensity, and grid flexibility are becoming binding constraints on AI compute growth [TechCrunch, March 2024] [Sustainable Tech Partner, March 2024]. That outcome is reachable, not merely aspirational, because the public record shows three ingredients already in place: a parentage in Sidewalk Infrastructure Partners, whose anchor investors include Alphabet, Ontario Teachers’ Pension Plan, and StepStone Group [Verrus, retrieved 2024]; a product thesis centered on integrating onsite battery storage with proprietary energy-management software [Verrus, retrieved 2024] [TechCrunch, March 2024]; and an external technical validation path through NREL’s ARIES Virtual Emulation Environment and the 70 MW Vulcan test platform [OSTI.GOV, retrieved 2026] [Latitude Media, retrieved 2026]. For a category where customers care as much about power certainty as they do about white space, that combination matters.

The upside is easiest to see through a few distinct scale paths rather than a single all-or-nothing case.

Scenario	What happens	Catalyst	Why it's plausible
Grid-constrained AI campus developer	Verrus becomes the preferred developer for hyperscale AI workloads in markets where utility interconnection is the main bottleneck, using battery-backed designs to make projects financeable and operable sooner	A first commercial deployment in one of the publicly discussed geographies, Arizona, California, or Massachusetts, demonstrates that integrated storage can secure capacity where conventional builds face delays [TechCrunch, March 2024]	Verrus has already framed those geographies publicly and has positioned its architecture around balancing power, carbon, and compute requirements [TechCrunch, March 2024] [Verrus, retrieved 2024]
Flexible compute infrastructure operator	Verrus wins customers with variable-uptime AI and batch workloads by offering high availability when required and optional flexibility when workloads can be shifted	A named customer or partner validates the model that some compute can tolerate workload shedding during peak system strain in exchange for access to scarce capacity [Verrus, retrieved 2024]	The company has publicly described differentiated service characteristics for flexible AI training and batch computing, and trade coverage has repeated the grid-aware and compute-aware positioning [Verrus, retrieved 2024] [Sustainable Tech Partner, March 2024]
Energy-software-led infrastructure platform	Verrus uses the energy management layer as a repeatable control plane across multiple sites, turning one successful campus into a multi-site development program with utility and power-market relevance	Commercial follow-through from the NREL demonstration shows that the operating logic works under real-world grid conditions, not only in concept [OSTI.GOV, retrieved 2026] [Latitude Media, retrieved 2026]	The public evidence already points to a technical demonstration on a 70 MW test environment, which is more substantive than a slideware claim and suggests the software layer is central to the thesis [OSTI.GOV, retrieved 2026] [Latitude Media, retrieved 2026]

The compounding mechanism, if it emerges, sits at the junction of infrastructure development and operating software. A first site that proves battery-backed flexibility can support both customer uptime and grid interaction should make the second site easier to permit, finance, and sell, because the proof point is not just demand for racks but demonstrated control over power quality, carbon procurement, and workload balancing [Verrus, retrieved 2024] [Sustainable Tech Partner, March 2024]. The NREL work is relevant here because it suggests Verrus is trying to validate a repeatable operating model, not merely a one-off engineering configuration [OSTI.GOV, retrieved 2026] [Latitude Media, retrieved 2026].

There is also a quieter form of lock-in embedded in the concept. If customers increasingly buy compute capacity on the basis of power availability and carbon profile, then a data center operator that can coordinate energy storage, procurement, and workload flexibility could have an advantage that is harder to replicate than a standard colocation buildout [TechCrunch, March 2024] [Verrus, retrieved 2024]. The same system intelligence that helps win an initial deployment could improve siting, procurement, and commercial underwriting on the next one. That is the kind of operational flywheel investors usually want to see in physical infrastructure, even if the public record does not yet show it in commercial form.

The size of the win is large enough to justify close attention even with sparse public traction data. TechCrunch reported that a hyperscale Verrus facility could cost about $1 billion, with "hundreds of millions of dollars of equity" required [TechCrunch, March 2024]. If the company were to prove the model across several campuses and become a trusted developer-operator for grid-constrained AI capacity, the enterprise value could plausibly be measured in the multi-billion-dollar range (scenario, not a forecast), simply because each successful site would represent very large deployed asset value and potentially scarce capacity in a supply-limited market [TechCrunch, March 2024]. The public evidence does not yet support a tighter valuation bridge than that, but it does support the core point: this is a bet on whether a new infrastructure design becomes economically necessary as AI demand collides with grid limits.

Partially corroborated -- Based on one independent launch profile, one trade profile, one technical report, and company materials; the commercial upside case remains plausible but only partially corroborated in public sources.

Sources