Technology
as Resilience.

Reservoir VC invests at the pre-seed and seed stage in rugged AI and ag robotics. We take concentrated positions
in companies whose tech starts with one high-value task in ag 
and compounds to adjacent industries.

Introducing Rugged AI

Built for the hardest-to-fill and hardest-to-automate jobs.

Hardened, software-defined hardware that works in weather, terrain, and signal loss that would disable conventional systems.
01

Operates without connectivity

Perception and decision-making run entirely on-device — no cloud dependency.

02

Disruption is normal operation

Designed to handle weather, terrain, and biological variation as the default, not the exception.

03

Modular and transferable

The core stack moves across crops, machines, and industries without a rebuild.
Image
PORTFOLIO HARDWARE · BONSAI ROBOTICS
The thesis

Rugged AI originates in
agriculture — and the transfer
of capabilities

Three distinct technology assets mature on farms first. Each maps to the same problem in adjacent industries.
The environment

Rugged autonomy — field operations

Rugged autonomy fills the hardest jobs — and the hardest to fill. Let the robot bear the heat, hold the harsh positions, handle the toxic exposure.
Proven in ag
autonomous tractors & carriers running full seasons in row-crop and orchard conditions
Mining · off-road haul & inspection
Energy · remote site operations
Forestry & land mgmt  · terrain work, fire-fuel reduction

The object

Perception & manipulation — precision interaction

How you engage an individual object: see one irregular, deformable thing in outdoor light and clutter — then pick,
cut, or dose it without damage.

Proven in ag

apple harvest, vine pruning, plant-level targeted spray

Forestry · seedling handling, precision planting
Energy · live-line & confined-asset work
Mining · core & sample handling

The target

Treatment modality — chemical replacement

What you do to the target: light, heat, electricity as the mode of action. Efficacy is target-specific, not crop-
specific.

Proven in ag

UV-C pest control, steam soil treatment, electric & laser
weeding

Utilities & rail · herbicide-free vegetation mgmt
BLM rangeland · invasive seed bank destruction
Nurseries & aquaculture · pathogen control

Why agriculture first

Agriculture builds the grit that
rugged AI requires.

Agriculture isn't the easy market, but rather the proving ground. The hardest problems live here, and what survives the field doesn't stay in the field.

01

Generational scale

$1.5T under pressure. The conditions that create category leaders.

02

The hardest problems live here

Unstructured terrain. Biological variability. Real-world autonomy.

03

Built to transfer

Built for row crops. The same AI stack powers mining, forestry, and energy.

Case in point

Chemical replacement

Precision targeting — laser, UV, mechanical weeding — is replacing herbicides in specialty crops today. The same see-and-act capability becomes vegetation management for utilities, fuel-load reduction for wildfire, selective treatment in forestry. One problem, solved in the field. Deployed everywhere.
Image

Solved here

Agriculture

Precision & non-chemical
weeding

Transfers to

Utilities & rail

Right-of-way vegetation
management

Transfers to

Forestry

Wildfire fuel-load reduction

Transfers to

Land mgmt

Selective, chemical-free
treatment

Why now

The rugged AI stack just matured.

Startups used to build everything custom. Now the infrastructure exists — SLAM, edge compute, simulation — so founders compete on application, not plumbing. This is the window.

70%+

of farmland has no reliable signal —
now addressable

$12.2B

agricultural robotics market, growing
~25% CAGR

1 yr

of foundational perception work
startups now skip

Image

The window opens

2018
Custom everything, high burn

2020

ROS 2 matures, shared tooling

2022

Edge compute becomes cost-viable

2024

Visual SLAM & simulation ship

Now

Visual SLAM & simulation ship

What we back

A focused fund for early-stage
rugged AI and ag robotics.

The fund at a glance

Pre-seed → Series A

Stage we invest

rugged AI & ag robotics

~$750K

Average check

per company

What kind of company

Application-led

Ruggedized robotics

The robots doing the work — non-chemical weeding, harvest automation for the hardest-to-fill roles, plant-level inputting, and field monitoring that turns sensor data into action.

$12.2B

ag robotics market, 2025

Capability-led

The rugged AI stack

The intelligence powering them — outdoor perception and real-world models, task-level autonomy that works without connectivity, fleet learning, and simulation that adapts across crops and geographies.

$7.3B → $19B

robotic software platforms by 2030

compound opportunity of rugged ai

TRIC Robotics

~40× TAM

UV pest control for strawberries → multi-task crop protection

TerraBlaster

10-12x TAM

Precision soil treatment robotics — backed by Khosla Ventures post-Reservoir

Bonsai Robotics

$6B+ adjacency

Single-machine autonomy → multi-OEM, now entering mining

RESERVOIR FARMS · VALIDATED BY WESTERN GROWERS

How we help founders

Reservoir Farms – Our own proving ground.

Three working farms across California's top specialty-crop regions, where portfolio startups test in real conditions — backed by a John Deere dealer network, university research, and $8M+ in non-dilutive capital.

3

Working farms in california

100+

acres of active testing

01

Top Shop & Crops

Startups can use on-farm machinery shop and field test all within steps of eachother

02

Embedded Grower Network

Startups gain access to an ecosystem of growers ready to adopt new tech

03

Industry Access

We put your tech in front of the eyes of John Deere and access to NVIDIA’s inception program
VISIT RESERVOIR FARMS

The team

We are investors, operators, and field experts.

Rare depth across technology, agriculture, and venture.

Image

Danny Bernstein

Managing Partner

Exit to Google ($100M). Former Managing
Director at Google and Partner at Microsoft.
Stanford GSB · UC Davis.

Image

Nicola Kerslake, CFA

General Partner

$240M+ raised in VC, 25%+ IRR on last fund. 2x
exited AgTech founder (Contain Inc., Indoor Ag-
Con). Prior PE at SEI Investments ($160B AUM).

Image

Matt Hoffman, PhD

General Partner

Led AgTech at Driscoll's — the world's largest
berry company. Deep field expertise across
viticulture and specialty crops. PhD, UC Davis.

Next steps

Get in touch.

01

"*" indicates required fields

Accepted file types: pdf, ppt, pptx, docx, gdoc, Max. file size: 10 MB.
Accepted file types: mp4, mov, Max. file size: 100 MB.