🌾 YOU COMMENTED "LAND"

the agritech
+ land intelligence hub.

every link here is a real, live resource. the jobs at the AI-meets-soil intersection, the repos to clone, and the open datasets a world model for land actually runs on.

THE JOBS THAT AREN'T BORING

the sweet spot isn't "agriculture jobs." it's jobs where land is the domain and the skill is expensive — geospatial ML, remote sensing, crop modeling. rough US ranges, directional not gospel.

🛰 agricultural remote sensing

$120–220K

remote sensing scientist · earth observation scientist · geospatial ML engineer
Regrow · Boomitra · MapMyCrop

🌱 soil carbon MRV

$100–200K+

soil carbon scientist · MRV scientist · carbon data scientist
Boomitra · Varaha · Agreena

🌍 land intelligence / geo AI

$120–220K

geospatial data scientist · land intelligence scientist · spatial ML engineer
MapMyCrop · EarthOptics · Overstory

🤖 agricultural robotics

$140–250K+

perception engineer · field robotics scientist
Carbon Robotics · FarmWise · Bluewhite

🧬 computational plant science

$120–220K

plant data scientist · crop modeling scientist · computational biologist
Pairwise · Inari · Benson Hill

🌦 climate × ag modeling

$110–200K

crop modeler · agroclimate scientist · climate risk scientist
Tomorrow.io · Climate Corp · EarthDaily

💧 water intelligence

$100–190K

agricultural hydrologist · irrigation data scientist
Kilimo · Waterplan

🌾 regenerative ag intelligence

$90–180K

regen ag scientist · practice verification scientist
Varaha · Regrow · Nori

🛰 crop insurance intelligence

$100–200K

agricultural risk scientist · geospatial risk scientist
Ceres Imaging · Descartes · Arbol

🌳 nature / ecosystem intelligence

$100–200K

ecological data scientist · nature MRV scientist · biodiversity analyst
NCX · NatureMetrics · Pachama

the positioning that opens the most doors isn't "agtech person." it's geospatial AI → physical-world intelligence → agriculture. that same stack works for climate tech, insurance, carbon markets and satellite companies too.
REPOS TO CLONE

flagship, actively-maintained repos — the real thing, not hello-world notebooks. study these, then fork one.

🛠 geospatial ML core

🌾 crop & plant

the modern shift worth learning: instead of training a CNN on Sentinel-2 from scratch, use a pretrained EO foundation model (Prithvi, SatMAE, Clay, DOFA) → extract embeddings → fine-tune for crop / yield / soil. TorchGeo ships several of these.
THE DATA GOLDMINE

free, planetary datasets — imagery, crop maps, soil, weather. this is the raw material. bookmark all of it.

🛰 imagery

🌾 crop & yield

🌱 soil

🌦 weather & climate

📚 knowledge / text

🧪 learn & build

PLANT, PEST & DISEASE VISION

the image datasets for the biology side — leaves, pests, disease. good for computer-vision portfolio projects.

🌿 disease

🐛 pests

  • IP102102 pest categories, fine-grained

🔬 phenotyping

  • PlantCVleaf area, morphology, growth
WHY I CARE ABOUT THIS

🌱 i'm building pestren.

a land intelligence engine. computer vision + knowledge graphs + survival models on one side — nine generations of knowing one piece of land on the other.

forward-deployed engineer + ninth-generation
farmland owner. i don't romanticize the soil —
i want to model it. that's the bet.

SEE PESTREN ↗ JOIN THE WAITLIST ↗
one honest note on licenses: "publicly downloadable" ≠ "free to train a commercial model on." before anything goes in a training set, check the license, attribution, and whether it holds farmer/field-level data. government sources have terms too, even when access is free.

made with ♥ + a lot of soil
reply and tell me what you're building — i read every one