Go to any energy conference, or listen to any energy podcast, and the conversation is likely the same: data centers need power, the grid can't keep up, and we need to build faster. And for good reason, the AI boom has made this the defining infrastructure and energy challenge of the day.

But data centers, for all their urgency, still remain a fraction of overall electricity demand. 

Today, they consume roughly 4.5% of U.S. electricity. Even in EPRI's most aggressive 2030 forecast, that figure reaches 17%, leaving the majority of current and future U.S. power demand potentially underserved as we (investors and innovators) collectively spend our attention and investment dollars on data center-specific solutions.

Non-data center commercial and industrial (C&I) loads actually make up a majority of US power demand. These are the large and medium-sized loads across manufacturing, smelting, food and beverage, hospitals, and chemical plants that consume over 55% of total U.S. electricity today.

This piece focuses on traditional C&I loads, like those mentioned above, and provides insight into the startups innovating in this space. 

Unprecedented grid congestion driven by AI infrastructure build-out, industrial electrification, increasing power quality disruptions and regulation, and a volatile U.S. energy market have handed C&I load operators a new risk profile that requires a swath of solutions tailored to their unique operational constraints. We’ll explore some of these pain points and the emerging solutions to meet this moment.  

If you’re a seed-stage founder developing digital solutions for grid innovation, or if you’ve been looking at C&I energy innovation, we’d love to hear from you. And if you know someone building in this space, please tag them in the comments. 

TL;DR (Tremendous Loads; Demand Rising)

• Volatile energy markets are creating new risks and opportunities for C&I loads: Increasing turbulence in capacity and demand charges is driving up costs as well as opening new monetization pathways. Most C&I operators face these dynamics with limited flexibility and few tailored solutions, but this is changing. 

• Power quality is becoming an operational risk: Dynamic loads and aging infrastructure are driving more frequent power quality-related grid disturbances, with $60B in annual industrial losses and limited real-time control. Customer- and utility-side monitoring and control platforms are emerging to close this visibility gap, giving operators real-time insight into disturbances before they cause damage.

• C&I flexibility is real but under-aggregated: The DOE estimates flexible C&I load potential as high as 300 GW today, and projects it could supply nearly half of all cost-effective demand flexibility nationally by 2030, but demand response aggregators typically haven't allowed operators to retain site-level control. Furthermore, to effectively self-dispatch into independent system operator (ISO) markets typically requires energy market expertise that most operators lack. Site-tailored flexibility platforms are starting to aggregate these complex loads, offering operators tailored levers of control with the right amount of automation. 

• Grid capacity for new loads exists but is difficult to find or utilize: Across transmission, distribution, and behind-the-meter systems, a need for better visibility and coordination leaves usable capacity stranded. The opportunity to monetize on-site surplus capacity is emerging.

Pain Point 1: New Challenges in Maintaining Power Quality

Power quality refers to how consistently electricity is delivered within required voltage, frequency, and waveform limits so that equipment operates safely and without disruption. Historically, utilities have been responsible for maintaining power quality, relying on grid-side hardware such as capacitors, voltage regulators, harmonic filters, and transformers to manage relatively stable and predictable load profiles. 

However, the proliferation of dynamic, nonlinear loads is creating new challenges for both utilities and C&I customers like manufacturing plants, hospitals, warehouses, and commercial buildings that never before thought of themselves as grid-edge infrastructure. Industrial facilities experience an average of 66 voltage disturbance events per year, contributing to a conservative estimate of $60B in annual losses from resulting operational downtime, on-site equipment damage or resets, and reduced throughput from slower cycles times and rework. Surprisingly, up to 80% of disturbances originate within facilities, rather than from the grid, yet utilities lack visibility into behind-the-meter behavior, and C&I operators lack real-time tools to manage power quality across complex systems.

This growing mismatch between increasingly dynamic loads and static, siloed power quality solutions is creating new opportunities for technologies that provide real-time visibility, faster responses, and coordinated control across both sides of the meter.

Opportunity: Putting power quality control in customers' hands

These solutions shift power quality from a passive protection problem to an actively managed, real-time control layer at the site. Deployed and operated by the customer, they combine hardware and software to dynamically condition power, manage disturbances, and optimize interactions with the grid.

• Real-time power conditioning and control: Power electronics, including solid-state transformers and advanced inverters, enable sub-cycle response to voltage and frequency deviations while actively managing reactive power and harmonics.

• Grid-forming and fast-response assets: Behind-the-meter (BTM) storage is evolving from a resiliency tool into a power quality buffer, capable of providing fast frequency response, synthetic inertia, and voltage support.

• Hybrid storage systems: Pairing batteries with supercapacitors and management software enables response across multiple timescales, from milliseconds to minutes, addressing both transient spikes and sustained events.

• Power quality orchestration software: Emerging platforms coordinate multiple on-site assets, such as UPS, storage, and filters, into a unified control system rather than isolated point solutions. Integrating energy controls with industrial control systems allows loads to dynamically adjust operations in response to power quality events without disrupting production.

Startups: Relyion, Lyra Power Technologies, Florrent, Heron Power, DG Matrix, Powerside, VoltVision, Energy Quotient (all startups mentioned will be listed and linked below)

Opportunity: Give utilities better power quality tools

While utilities have strong visibility at the transmission level, power quality challenges are increasingly emerging at the grid edge. This creates an opportunity to extend sensing, analytics, and control deeper into the distribution network.

• Grid-edge observability and high-resolution sensing: Expanding monitoring beyond substations using smart meters, line sensors, and edge devices to capture voltage, harmonics, and transient events in real time.

• High-frequency data and event detection: Moving from interval-based data to waveform-level monitoring enables detection of sub-second disturbances that traditional systems miss.

• Predictive grid-edge power quality management: Platforms that combine advanced metering infrastructure (AMI) data, grid modeling, and controllable distributed energy resources (DERs) to monitor voltage quality, detect anomalies, anticipate disturbances, and proactively support local voltage regulation and stability.

Startups: Utilidata, Grid4C (acquired by Bidgely), PingThings, Reactive

Pain Point 2: Complex C&I Loads Are Hard to Aggregate  

Large C&I loads are often held up as a potentially cost-effective grid flexibility resource. The DOE estimates flexible C&I load potential as high as 300 GW today, and projects it could supply nearly half of all cost-effective demand flexibility nationally by 2030. However, realizing this resource has been challenging. Large sites across commodities like chemical, metals, food and beverage processing often operate heterogeneous, uptime-critical processes that cannot be easily curtailed without impacting production, making them difficult to integrate into existing demand response programs. This has led aggregators like CPower and Voltus to lean on semi-standard and easy-to-dispatch loads because managing more complex and mission-critical loads to date has been cost-prohibitive. With new tools enabled by the step change in AI technology, this currently underserved pool of complex C&I loads could finally become an aggregatable, controllable flexibility resource.

Opportunity: Building site-tailored flexibility tools for complex loads

• Process-aware flexibility modeling and real-time load orchestration: Software that maps industrial processes and constraints to identify where flexibility exists, then coordinates multiple behind-the-meter assets for automated, sub-hour/minute dispatch with guardrails that include a customer's specific operational parameters. 

• Industrial-native VPP participation layers: Tools that abstract market complexity and enable C&I operators to participate in demand response, capacity, and ancillary markets without needing in-house trading or regulatory expertise.

• Portfolio-level flexibility aggregation: Systems that aggregate flexibility and unify existing energy contracts and applicable tariffs across multiple sites are critical to enabling operators to optimize participation and revenue across their full footprint rather than on a site-by-site basis.

Startups: Companion Energy, GridBeyond, Encentive, Sympower, Snerpa Power

Pain Point 3: C&I Loads Need Tools to Access or Expand Capacity  

At the transmission level, where large facilities of 20 MW and above interconnect, the queue has become saturated with data center demand, crowding traditional industrial load requests that, in comparison, typically have thinner margins, fixed production timelines, and limited means to self-fund their way to the front of the interconnection queue or directly execute costly network upgrades to secure access.

At the distribution and sub-transmission levels, where most brownfield expansions and mid-sized C&I projects connect, capacity is becoming increasingly opaque and fragmented. Distribution networks remain predominantly analog and lack real-time monitoring capabilities. This makes it difficult for utilities to determine where capacity actually exists. Even incremental energization of existing sites is slowed by fragmented regulatory processes and growing queues, creating friction for multi-site operators trying to scale or comply with reshoring mandates. The net effect is a system where capacity may physically exist, but is difficult to identify or utilize efficiently, leaving manufacturers, hospitals, and industrial sites without the tools or transparency needed to plan growth with confidence.

Opportunity: New monetization strategies for stranded grid capacity at the site 

The same constraints that make new capacity difficult to access are increasing the value of existing, underutilized connections. Many traditional C&I sites are provisioned for peak demand but operate far below that level, leaving significant unused capacity and the opportunity to build new revenue streams. 

• Capacity monetization platforms: Enabling existing C&I operators to lease or sell their excess capacity to third parties without new interconnection

• Colocation and energy arbitrage models: Pairing new loads, namely distributed data centers, with existing industrial sites to bypass queues and reduce time to power

• Capacity utilization analytics: Identifying and quantifying unused headroom across portfolios to unlock new revenue streams

• Regulatory and interconnection structuring: Navigating tariffs and rules to enable capacity sharing while maintaining compliance

Startups: Green Weaver, Perimeter Compute, InfraSale 

Opportunity: Transmission and distribution-level capacity intelligence and access 

Like the flurry of data center-specific capacity intelligence platforms built to speed hyperscale interconnection, non-AI C&I loads need solutions tailored to their own transmission and distribution-level operational requirements.

• Faster study and interconnection intelligence: Software that simulates outcomes of interconnection studies or predicts upgrade requirements before entering the queue reduces risk and time to power.

• Capacity mapping and siting tools: Identifying “pockets” of available transmission capacity allows C&I developers to site projects more strategically.

• Feeder-level sensing and monitoring: Low-cost sensors and edge devices provide continuous visibility into voltage, load, and thermal constraints beyond the substation.

• Planning and operations integration: Tools that connect distribution planning with real-time operations to enable faster approvals and reduce conservative assumptions.

Startups: ThinkLabs and Pearl Street Technologies (Powerhouse Ventures’ Portfolio Companies), GridCare, Nira Energy, LineVision, Splight, Piq Energy, Camus Energy, Gridaptic 

Opportunity: Expanding C&I BTM capacity in this market moment

For many C&I operators, the time has come for real investments in BTM assets. Voltus’ recent acquisition of Brightfield AI marks a strategic shift away from decades of slow BTM growth. Battery hardware costs have fallen by a third in the past 18 months, and business models that once didn't pencil finally do.

• On-site generation and storage optimization: Solar, storage, and hybrid systems used not just for cost savings but to increase effective site capacity

• Load shaping and peak management: Software that reshapes demand profiles to avoid triggering costly upgrades or demand charges

• Procurement and financing platforms: Streamlining deployment of BTM assets by reducing upfront cost and complexity

Startups: Verse, Branch Energy, AiGent, Box Power, Brightfield AI (acquired by Voltus), Edgy Systems

Powerhouse Ventures Perspective: Unlocking the C&I Energy Stack 

For decades, many commercial and industrial facilities and energy operators treated electricity as a relatively predictable line item. While some opted to be part of standard VPP aggregation programs or invest in their own on-site energy systems, the majority weren’t a fit or didn't feel real pressure to participate. We’re starting to see this dynamic shift.

It’s worth noting that several of the startups listed above target data centers as their entry customer with the intentionality to serve other commercial and industrial load types over time. We hope this article highlights the need for more innovation for medium- and large-sized C&I sites that have a pressing and complex need for solutions today given the rapidly shifting market dynamics outlined here. 

We are excited to continue exploring and engaging with founders building solutions tailored to the realities and opportunities of the non-data center C&I segment. 

As a reminder, if you’re a seed-stage founder developing digital solutions for grid innovation, or if you’ve been looking at C&I energy innovation, we’d love to hear from you. And if you know someone building in this space, please tag them in the comments. 

Appendix: Companies and startups mentioned 

• Relyion - AI platform smoothing power draw with batteries

• Lyra Power Technologies - The dynamic interface between facilities and the grid

• Florrent - Ultracapacitors that help stabilize electrical grids

• Heron Power - Solid-state transformers built for the modern grid

• DG Matrix - Multi-port solid-state transformer solution

• Powerside - Power quality measurement, correction, and monitoring solutions

• VoltVision - AI monitoring and analytics for industrial power networks

• Energy Quotient - High-fidelity sensors for real-time industrial power quality monitoring

• Utilidata - Dynamic power orchestration platform

• Grid4C (Bidgely) - Combined AI platform for usage disaggregation and grid analytics

• PingThings - Nanosecond-precise AI platform for grid sensor data

• Reactive - Measures grid inertia and system strength in real time

• Companion Energy - Software managing industrial energy flexibility

• GridBeyond - Platform for managing distributed and flexible energy resources

• Encentive - Lets industrial companies get paid for load flexibility

• Sympower - Connects energy assets to the grid for demand-side revenue

• Snerpa Power - Software helping power-intensive industries optimize electricity consumption

• Green Weaver - Distributed AI compute infrastructure at industrial sites

• Perimeter Compute - Turns underused C&I space into GPU hubs

• InfraSale - Marketplace for powered land

• ThinkLabs - AI digital twins and agents for utilities and grid operators

• Pearl Street Technologies - Modeling tools that accelerate grid interconnection

• GridCare - Unlocking existing grid capacity

• Nira Energy - Transmission insights throughout the interconnection process

• LineVision - Sensors that increase transmission line capacity

• Splight - Dynamic congestion management software

• Piq Energy - Automates grid and interconnection studies

• Camus Energy - Gives utilities situational awareness across grid and customer data

• Gridaptic - AI analytics for utility distribution capital planning

• Verse - Platform accelerating data center interconnection via on-site assets

• Branch Energy - Commercial battery installs cutting costs and outage risk

• AiGent - Turns backup generators into revenue-generating, monitored assets

• Box Power - Turnkey microgrids integrating solar, battery, and backup generation

• Brightfield AI (Voltus) - Automating C&I battery storage projects

• Edgy Systems - Designs and operates microgrids for commercial and industrial sites