The grid isn't out of room. It's out of intelligence - and Splight ships the upgrade.
It is a summer afternoon and somewhere a grid operator is staring at a number that is lying to them. A transmission line rated for a fixed, conservative limit is actually running cooler than the rulebook assumes. There is real, usable headroom on that wire right now - capacity that exists, paid for, strung between towers that are already there - and the system is throwing it away. Multiply that by a continent of lines and you get the grid's quiet scandal: we keep talking about building more, while wasting what we already have.
Splight built its business in that gap. Where most of the energy world reaches for steel, permits and a ten-year wait, Splight reaches for software. Its premise is almost rude in its simplicity: the wires are fine. The grid just needs to think faster.
“Our technology solves one of the most urgent problems facing the grid today: how to deliver more electricity without waiting years for new transmission to be built.”
- Fernando Llaver, CEO & Co-Founder
Splight adds a real-time operational and reliability layer to the power grid. Sensors and models read what is actually happening on a line - temperature, weather, load, contingencies - and the system rates capacity dynamically instead of trusting a worst-case number set decades ago. Then comes the clever part: Splight turns fast-responding assets like batteries and data centers into instruments that can absorb shocks, so operators can run lines closer to their true limits and stay safe.
The flagship is Dynamic Congestion Management (DCM), a proprietary machine-learning system the company says can unlock up to 100% more usable capacity on existing transmission - no new lines, no new hardware. For an industry where the standard answer to congestion is “build,” that is a different sentence entirely.
Machine learning that safely unlocks up to 2X usable capacity on existing lines by responding to grid contingencies in real time.
Rates lines on real conditions, not conservative static limits - freeing the headroom the rulebook leaves on the table.
Faster interconnection and transmission studies for data centers and new loads, shrinking the queue to plug in.
A modeling tool to simulate congestion-management scenarios before anything touches the live grid.
Hardware-in-the-loop testing that validates control strategies against real equipment behavior first.
Existing line, static rating vs. Splight Dynamic Congestion Management
Figures are company-reported and vary by system characteristics. Approximate, illustrative.
Splight sells to utilities, transmission operators and the owners of utility-scale renewable and storage assets. But the more urgent customer is the AI data center. Hyperscale demand is arriving faster than transmission can be built, and a project that can't connect is just a spreadsheet. Splight's pitch to them is interconnection without the decade-long wait - capacity found, not poured.
The company reports more than 3 gigawatts of grid assets under management across multiple geographies and regulatory regimes, and is expanding across North America and Europe. Its business model is refreshingly unglamorous: B2B software layered on top of infrastructure that already exists, rather than a fortune in new construction.
Splight came out of the energy sector, not a Stanford whiteboard. Its founders spent time inside the industry's machinery - computer science, electrical engineering, and the unglamorous reality of running power systems - before deciding the smarter move was software.
Blue Bear Capital cited the team's deep operational knowledge of the grid and their machine-learning expertise - and a deployment track record that builds confidence with infrastructure operators.
- on the 2025 round, via Carolin Funk, Partner, Blue Bear Capital
Splight closes $12M to take on energy curtailment - the grid's habit of wasting clean power it can't move - and to push DCM into the field.
A SAFE round led by Blue Bear Capital with ZOMA Capital, aimed squarely at the transmission crunch created by AI data centers and utilities.
AI-enabled electrical studies for data centers, enhanced DCM algorithms, the DCM Sandbox modeling tool, and Real-Time Simulation for hardware-in-the-loop testing.
Grid assets under management pass 3 GW as the company expands across North America and Europe.
Demonstrated up to 2X usable transmission capacity at EPRI Incubatenergy Labs Pitch Day in Atlanta.
Grid assets under management across multiple geographies and regulatory environments.
Across seed and a 2025 SAFE round, from climate and infrastructure investors.
Back in that control room, the lying number is gone. The line carries what it can actually carry, the data center plugs in years sooner, and the clean power that used to be curtailed finally moves. Same wires. Same towers. A grid that finally knows how much it can do.
- Splight: intelligence is the new capacity