A farm robot that treats plants one at a time - and in doing so, quietly rewrites the economics of the field.
The SharpShooter's cameras see a weed the way a copy editor sees a typo - one flagged item at a time, at 480 per second.
For about a hundred years, the standard way to protect a field of carrots has been to spray the whole thing. Verdant Robotics thinks that is the bug, not the feature.
Here is the thing about spraying an entire field: most of the chemical never touches a weed. It lands on dirt, on the crop you are trying to grow, on the air. It is expensive, it is imprecise, and increasingly it is the kind of thing that regulators, retailers and shoppers ask pointed questions about. The traditional alternative - paying crews to walk the rows and pull weeds by hand - is expensive in a different way and, in much of California agriculture, getting harder to staff every season.
Verdant Robotics, founded in 2018 and based in Hayward, California, sells a third option. Its flagship product, the SharpShooter, is an implement you pull behind a tractor. As it moves, stereo cameras feed a bank of onboard AI that identifies every plant beneath it - crop here, weed there - and a set of aiming nozzles that fire a targeted dose at each one. Weeds get herbicide. Crops get fertilizer or protection. Nothing gets a blanket treatment.
The company calls the underlying trick "Bullseye Aim & Apply," and the marketing line is that it is the only precision system that aims before it applies. The engineering claim behind the slogan is more interesting than the slogan: 99% of shots land within five millimeters of their target, at a rate of 120 to 480 shots per second, on a machine bouncing across an open dirt field at up to five acres an hour. Doing that reliably, all day, in dust and heat, is the actual hard part - and it is roughly why the founders' résumés matter.
The results Verdant reports from customer fields are the sort of numbers that make a grower's accountant sit up: chemical inputs down more than 96%, hand-weeding costs down an average of 65%, and a payback period of 12 to 24 months. That last figure is the one that turns a nice-sounding sustainability story into a purchase order. When the greener option is also the cheaper option, you do not have to win a moral argument. You just have to do the math.
The founding team came to agriculture from some of the least agricultural places imaginable - which turns out to be the point.
A roboticist who previously built perception software for NASA rovers and self-driving cars. His recurring lesson for ag robots: they "have to be insanely robust," because a field is less forgiving than a lab.
Brings the commercial and grower-facing side, translating a very technical product into per-acre outcomes that farms will actually buy.
Runs operations, and frames the mission plainly: "We wanted to do something meaningful to help the environment" - while cutting cost and lifting yield.
They have to be insanely robust.
The SharpShooter is unusual because it does not specialize. In a single trip down a row it can weed, thin, fertilize and treat for pests - deciding, plant by plant, which action each one needs.
A 20-foot autonomous implement (the 2025 Model B) that treats individual plants at up to 5 acres per hour, handling targets from the size of a dime to a dinner plate.
Spatial AI plus aiming nozzles that lock onto a target before applying. 99% of shots land within 5mm, at 120-480 shots per second.
Growers can buy weeded, thinned and treated acres rather than the robot itself - Verdant operates the fleet and delivers the outcome per acre.
Every pass doubles as a scouting pass. The robot maps and diagnoses each plant, handing growers real-time, plant-level field data.
Reported results on serviced specialty-crop fields
A lot of companies can build a slick demo of a robot spotting a weed. The difference between that and a machine a farmer runs all season is a stack, not a single breakthrough. Verdant runs NVIDIA Jetson AGX Orin modules right in the tractor cab, doing real-time inference at 29 frames per second and rendering live 3D visualizations so the operator can watch each plant get treated. The detection and segmentation lean on NVIDIA's DeepStream SDK, with custom CUDA code for tracking, and models trained on DGX systems with A100 GPUs.
The quietly important number is this one: Verdant cut the time to retrain its crop AI from carrots to onions from five days to one, using transfer learning. In a business where every crop - and sometimes every field - looks a little different, the speed at which you can teach the robot a new plant is not a footnote. It is the moat. The company has already stretched from carrots, lettuce and onions into apples, garlic, broccoli, and even grass seed and sod, where the weeds and the crop can look nearly identical.
Sibley, Garner and Ibarria start Verdant Robotics in the Bay Area to bring precision robotics to farming.
Backing from AgFunder, Moog and DCVC funds SharpShooter development and the first crop trials.
Verdant begins servicing specialty-crop acreage commercially on a per-acre basis.
Cleveland Avenue leads one of the largest ag-robotics rounds to date to scale the fleet.
The 2025 Model B debuts at FIRA USA with Bullseye Aim & Apply, 5 acres/hour and 5mm accuracy.
The 2022 round was, at the time, one of the largest single investments in agricultural robotics - a bet that precision application is where the field is heading.
| Round | Amount | Date | Lead & notable investors |
|---|---|---|---|
| Series A | $46.5M | Nov 2022 | Cleveland Avenue (lead), DCVC Bio, Future Ventures, SeaX Ventures, Autotech Ventures, Cavallo Ventures, AgFunder |
| Seed / earlier | ~$11.5M | 2020 | AgFunder, Moog, DCVC |
| Total reported | ~$58M | — | Across rounds |
The SharpShooter delivers unmatched speed, efficiency, and versatility, giving growers the savings, precision, and control they need.
Verdant aims squarely at high-value, labor-intensive specialty crops - the kind where hand-weeding is expensive and margins reward precision. It is active on thousands of acres across California's Central and Imperial Valleys and Oregon, and its named work includes Bolthouse Farms, the Bakersfield carrot giant, which is using Verdant to move toward regenerative practices, starting with weeding and expanding into precision fertilizer.
Transitioning its carrot business toward regenerative agriculture - reducing herbicides, starting with robotic weeding.
Jetson in the cab, DeepStream and TAO for the AI, DGX with A100s for training the crop models.
Early strategic investor with motion and actuation expertise relevant to precision hardware.