I came up in this industry as an operator, not an investor. I spent the early years of my career at SolarCity and Sunrun when residential solar was a ten-dollar-per-watt product, and then I joined the founding team at Aurora Solar to automate the process and lower soft costs.

After “cleantech 1.0,” I worked on electricity market design and helped large corporates with sustainability strategy at the Rocky Mountain Institute (RMI). My background shapes how I invest.

I look for bottlenecks in our systems, then back the founders who can clear them with step-change value propositions.

It's no secret that the grid has become a major bottleneck to economic expansion. In fact, given the grid’s complexity, there are multiple bottlenecks across voltage levels, interconnects, and operations.

After decades of flat demand, load growth is back, driven by data centers, electrification and a manufacturing base coming home. But “bottlenecks" almost undersell what's happening.

The system is woefully congested and lacks streamlined orchestration and operations, which is showing up in the red all over the place. Generation isn't the shortage; at the end of 2025, roughly 2,000 gigawatts of generation and storage, fossil and renewables alike, sat waiting in interconnection queues, according to Lawrence Berkeley National Laboratory, and the median project that reached commercial operation last year had waited more than five years to get there.

Meanwhile, transformer lead times run in seasons, peak forecasts get revised upward every planning cycle, and data centers are gobbling up whatever spare capacity exists. The constraint is no longer making electrons; it is making them available and moving them to where they are needed via distributed orchestration, load shifting, and efficient forecasting and operations.

Risk fills the other side of the ledger. Wildfires stopped being a California problem years ago. The largest fire in Texas history, the 2024 Smokehouse Creek fire, started when a decayed utility pole snapped in high winds, and drier seasons keep turning more of the service territory map into fire country.

Storms keep setting records too, as Houston can attest after a derecho and Hurricane Beryl knocked out power to millions of CenterPoint customers within two months of each other in 2024. And not every threat arrives with the weather.

Physical attacks on substations, like the gunfire that blacked out 45,000 customers in Moore County, North Carolina, have risen sharply in Department of Energy incident data, and utility networks get probed constantly. Grid security and community safety have become board-level topics.

All of this lands on infrastructure built for a different era and never designed for what we are asking of it now, and all of it prices into insurance, liability, and capital plans. Investors read that as urgent demand for hardening, monitoring, and security.

On the supply side, the economics shifted around 2020, when wind, solar and batteries crossed cost thresholds that made them competitive with fossil generation, subsidized or unsubsidized, depending on where you look. Since then, the argument has only gotten stronger.

Subsidies are rolling back, and that's fine. These are mature technologies now, and technologies keep riding cost curves down in a way commodities never will. Every year that passes, the gap widens.

And with battery costs continuing rapid decline as well, the renewable “intermittency” argument is on its deathbed.  

The proof points are on your grid already. In April 2024, for the first time, batteries were the single largest source of supply on the California grid during the evening peak, and by that October they were covering more than a fifth of system demand at peak during a late-season heat wave, according to CAISO.

That answered a question the industry had been asking for a decade about whether storage could balance a demand curve at scale. Texas tells the same story from a different angle, where the wind and solar built across the western half of the state bids into ERCOT's day-ahead nodal market and pulls wholesale costs down.

None of this maps to politics. Deployment is happening in red states and blue states alike, because the value is there.

Here is the part I find most compelling as an investor, and it is something utility engineers have known forever. We size generation to meet a peak that lasts a few hours a day, which leaves an enormous amount of capacity sitting idle most of the time, waiting to ramp.

If we get serious about demand flexibility and storage, or load-shifting air conditioning technologies like Blue Frontier, we unlock latent headroom at GW scale from the existing system, which is exactly what we need as data centers line up for interconnection. Often the cheapest new capacity comes from getting more out of the steel we already have, rather than putting new steel in the ground.

Venture capital learned some expensive lessons between 2021 and 2023. Too many dollars went into solutions that needed hundreds of millions just to reach the market, and many still have not. 

The discipline that emerged from that cycle guides us now: back solutions with small to medium capital intensity, a step-change value proposition for the customer, and a fast path to infrastructure-scale capital. Applied to the grid, that points to the software and orchestration layer, and to hardware that pays for itself from day one.

Interconnection is a good example. Pearl Street Technologies built software that automates interconnection studies for ISOs and developers, collapsing timelines that had ballooned into multi-year backlogs.

It solved a bottleneck the industry felt every day, won real customers on both sides of the queue, and was acquired in 2025.

But there is a harder problem emerging as this buildout accelerates. Every fix arrives as its own system.

Consider dynamic line ratings to find more capacity on existing conductors, sensors that flag a failing transformer on Main Street before it fails, or monitoring for PSPS conditions and downed lines in dry fields before they become catastrophic fires. Each is valuable on its own, and each shows up as another dashboard, another data feed, another vendor login.

Nobody can operate a grid efficiently using myriad disparate, siloed systems. Utilities need a way to see it all in one place.

That is the problem companies like Texture, in our portfolio, are working on: a data orchestration layer that ties DERs, sensors, and grid-edge devices into a single picture operators can actually run the system from.

The technologies that will define the next decade of the grid are the ones that make your system more efficient, more reliable, and more resilient while saving money, and they need utility partners willing to pilot with intent to scale. Treat flexibility as capacity in your planning processes.

Value the software that clears your queues, gives you visibility to the grid edge, and shortens the time between a fault and a response. And hold every vendor, including the ones we fund, to the same standard: show the economics.

I am an economist and a capitalist before I am anything else, and what draws capital to the grid is not sentiment. Markets reward better, cheaper, more resilient solutions, and after five years of proof points, the direction is clear.

The load is coming, the threats are real, and the tools finally pencil. What happens next depends on how fast we build.