Startup Raises $11.2M to Decode Plant Electricity With AI

Electric Plant has raised an $11.2 million seed round to bring an AI model called Bombadil to market — a large language model designed to read the electrical signals flowing through a plant's stem and tell growers what the plant actually needs. Onto Ventures and Climactic led the round, with participation from Ponderosa, Endeavor Catalyst, Ludlow, Michigan Rise, The Fund, Schox VC, and Dug Song.

The startup's co-founder and CEO, Justin Mast, came to the field through a single strange experiment. A few years ago, while reading scientific literature, he found research on a houseplant wired to a motorized stand. When the plant wanted more light, electrical signals moved through its living tissues, a control board sensed the current, and the stand wheeled the plant toward the light source.

"I'm like, what the heck?" Mast told TechCrunch. "That was the entry to the rabbit hole for me."

That rabbit hole ended in a company. Mast partnered with Lee von Kraus, a neuroscientist who has worked with DARPA, to found Electric Plant. The team has spent the past three years collecting data to build Bombadil, which can interpret the subtle electrical signals in roughly 125 different plant species.

Why would anyone build a plant language model?

The stakes come down to a blind spot in modern agriculture. Even in tightly controlled greenhouses, growers don't have full visibility into what their plants require, Mast said. Farmers, growers, and foresters have traditionally relied on coarse tools — soil probes, handheld spectroscopes, satellite imagery — that react to a plant's demands for water, nutrients, or light after the fact.

"That technology is sophisticated, but its ability to get at the inside of the plant isn't very sophisticated. Most of that stuff still lags quite a bit behind the real-time needs of the plant," Mast said.

Bombadil, by contrast, listens at the source: the plant itself. The founders believe their model could eventually support a tool that lets people "speak" with plants. "We're listening more than we're speaking, but listening's the first step," Mast said.

The company trained Bombadil using data from existing measurement devices — a starting point, not a ceiling. "We know the model will surpass a lot of that equipment," Mast said.

How do the sensors actually work?

The hardware side is deliberately cheap. Competing tools generally cost thousands of dollars per unit. Electric Plant's sensors cost "tens of dollars," according to Mast.

The details:

• Each sensor is about the size of a regular iPhone folded lengthwise.
• It attaches to the plant with stainless steel screws.
• The screws must penetrate the cambium — the layer of living tissue just under the bark — to pick up microvolt changes inside the plant.
• Data travels to a base station over a low-power wireless network.
• Electric Plant says it needs only a couple of sensors per acre for accurate results.

Once the data reaches the cloud, Bombadil interprets it and sends results back to the customer. A grower can then adjust watering or fertilizer schedules based on what the plant is signaling. Mast claims such a system can boost yields while cutting water use by 20% or more.

What can Bombadil understand today?

The model's first hundred-plus species spans a wide range of plant forms — ferns, redwoods, tomatoes, and dahlias among them, Mast said. That breadth is the company's technical calling card.

"This is the most diverse model ever created on biological electrical signals," Mast said. "No one's ever done anything equivalent to this in the animal world."

Mast would not disclose who Electric Plant's initial customers are. Agriculture and forestry are the obvious applications: perennial crops, timberland, and greenhouse operations all stand to gain from real-time insight into plant stress, water demand, and nutrient needs — sectors where water savings of 20% or more carry both economic and environmental weight.

What happens next?

The seed money gives Electric Plant runway to move Bombadil from research data to deployed hardware in the field. The company's bet is that listening directly to plants — at tens of dollars per sensor, a couple per acre — beats inferring their needs from soil chemistry and satellite views. If Bombadil holds up outside the greenhouse, growers may soon manage water and fertilizer by what the crop itself says, in the plant's own electrical language.