Honest signalsfor Prairie farmland.
Farm Credit Canada publishes land values once a year. Farmers, land investors, and brokers make six- and seven-figure decisions every week. AgriIntel closes that gap with a Tuesday-morning scorecard for every Small Area Data Region in Saskatchewan and Alberta.
The data was public. Nobody had put it together.
Statistics Canada has wheat yields by region going back to 1976. Sentinel-2 photographs every field on the Prairies every five days. ICE and CBOT price canola and wheat every trading day. Farm Credit Canada publishes land value baselines. All of it is free. All of it is in incompatible formats, on different schedules, from different agencies.
AgriIntel normalizes those sources into one weekly scorecard per region, so that a farmer in Yorkton or an investor in Calgary can see, in one place, whether their region is drifting from its historical baseline — and how much to trust that reading.
Signals, not advice.
- Flags regions where current wheat yield deviates sharply from the 10-year average
- Tracks multi-year yield trend, satellite vegetation, weather, and crop condition per region
- Reports a gross revenue-to-land-value ratio to compare regions on income potential
- Scores confidence by how much the indicators agree
- Says "no signal" when the data is missing, instead of inventing one
- Predict prices, yields, or weather
- Tell you to buy or sell anything
- Measure profitability — revenue yield excludes input costs, rent, and financing
- Value a specific parcel — signals are regional, not field-level
- Replace an agronomist, appraiser, or your own judgment
Confidence measures agreement, not size.
Every region gets a confidence score from 30% to 80%. It is not a probability and it is not about how large the signal is. It measures how many of the five underlying indicators — yield anomaly, yield trend, satellite NDVI, weather, and crop condition — point the same direction.
80% means they nearly all agree. 40% means they conflict and you should dig deeper before acting. 30% means the yield data itself is missing for that region, and everything else is context without an anchor.
The cap is 80% on purpose. A regional model built on public data should never claim certainty.
Institutional-grade, publicly verifiable.
Two of the fourteen planned metrics — land velocity and listing volume — are not yet live. They require a Saskatchewan land sales data license that will be funded from subscription revenue.
Built by someone who has spent a career making numbers traceable.
Close to twenty years in data and technology, most of it as a professional-services consultant to organisations in regulated, high-stakes industries — IBM, Suncor, Teck Resources, Chevron, Transport Canada — and in clinical research, building CDISC-compliant datasets for regulatory submissions. The common thread: systems where a wrong number has consequences, so every number has to be traceable.
Since 2018, Founder & CEO of NateAnalytics Solutions, a Calgary technology studio building software and IoT products for regulated, data-intensive industries. AgriIntel's confidence scoring, versioned formulas, and explicit "no signal" handling come straight out of that background. I handle the architecture and development directly.
AgriIntel has been publishing weekly since April 2026. If something looks wrong, reply to any digest — I read everything.
The sample digest shows the exact format subscribers receive.
View sample digest → or open the live dashboard