SE Ventures AI investments target power, grids and industrial automation

SE Ventures AI investments are moving further into the physical infrastructure behind AI, as Schneider Electric’s $1 billion venture arm looks at data centers, grid resilience, robotics and industrial automation. Amit Chaturvedy, the firm’s global head and managing partner, told Crunchbase News that the constraint around AI is no longer only chips or models, but the ability to secure buildings, power, electrification equipment and other capacity needed to build.

Schneider Electric, whose business spans energy management and automation, is using its venture arm to back startups tied to what Chaturvedy described as AI’s push into the physical world. No new fundraise or individual startup round was announced. The discussion instead points to where a large industrial incumbent believes venture returns may emerge as AI spending moves from software into capital-intensive infrastructure.

SE Ventures has eight unicorns in its portfolio, according to Crunchbase, and has recorded 12 exits. Its latest cited exit is Fabric8Labs, a 3D metal printing company acquired by Tokyo-based electronics manufacturer TDK Corp. Chaturvedy joined SE Ventures in 2022 after leading corporate investments at Cisco.

What is SE Ventures investing in around AI?

Chaturvedy named three broad areas: AI infrastructure, the power grid and industrial AI. In AI infrastructure, he pointed to portfolio company Together AI as an example tied to training and inference demand, and Hammerhead AI as a bet on data center efficiency that he expects to matter more once the current data center capital spending cycle slows.

For Schneider, the connection is commercial as well as financial. Chaturvedy said about 80% of SE Ventures portfolio companies have some form of commercial relationship with a Schneider Electric business unit. In many cases, he said, the startup works with Schneider as a partner serving customers, while vendor relationships make up a smaller share.

Grid-related software and hardware are another focus because AI data centers are adding to electricity demand that was already rising from electrification. Chaturvedy told Crunchbase News that new demand from AI is far larger than the load created by home electric-vehicle charging, and said startups that improve grid resilience are an obvious area of interest for SE Ventures.

He also described energy as a cost problem at the token level. Training and inference require electricity, so cheaper token production and lower token consumption are part of the near-term optimization work. In the middle layer, he cited shifting inference away from peak electricity periods and improving HVAC and cooling systems. Longer term, he said new generation capacity, renewables and battery energy storage systems are areas the firm watches closely.

Why does this matter for industrial startups?

SE Ventures is treating AI as a catalyst for industrial automation, especially where robotics and factory operations are constrained by labor and aging technical workforces. Chaturvedy said general-purpose models could let the same robotics hardware perform more varied tasks, a shift he linked to portfolio company Skild AI.

He also cited Axion, which works with warranty data and feeds analysis back to design engineers, as an example of AI applied to industrial workflows rather than broad software productivity. The claim from SE Ventures is that industrial customers will move beyond pilots when startups can produce measurable operating improvements. Specific revenue figures, adoption rates and portfolio company valuations were not disclosed.

Chaturvedy expects the transition to show up first where capital spending is already active, with data centers at the front. Over the next three to 10 years, he said more new industrial projects could be designed from the start around robotics and automation, as companies replace or expand older manufacturing capacity.

This story draws on original reporting from Crunchbase News.