Agronomy for all, on demand

From soil test to prescription, in one conversation.

Fieldhand AI talks through your soil test in plain language, while software tools compute every rate from published soil-science equations, so each number can be checked.

Create a free account →No account needed to ask a question.

fieldhand.aiterra.ag

My soil test shows phosphorus at 12.5 ppm by Bray-1. I'm growing corn for 210 bu/ac in Iowa. What should I do about P this season?

At 12.5 ppm Bray-1, your phosphorus is Low by Iowa State University guidelines (optimum is 16–20 ppm). Plan a build-and-maintain application this fall. I split the field into four zones by soil test, from 70 to 125 lb P₂O₅ per acre.

Iowa State PM 1688Calculated by tool · Iowa State PM 1688 table

Phosphorus prescriptionExample field · 80 ac · avg 107 lb/ac

Approve

Zone	Soil P	Rate
Zone 1	10.0	125
Zone 2	11.9	110
Zone 3	13.6	100
Zone 4	17.8	70
Field	12.5	107

Soil P in ppm, Bray-1. Rate in lb P₂O₅/ac.

Example conversation. Hover or tab to a zone to see its rate.

3issued US patents

AgLaunch365

Team from Uber · IBM · Google · Bayer · Land O'Lakes

The gap

Growers have the data. Few have the time or expertise to use it.

Sorting soil tests, yield maps and guidelines into a decision is skilled, tedious work.

1 in 3

corn and soy acres use variable-rate, though yield maps already cover 70–79% of those acres.¹ Dealers report no growth in variable-rate fertilizer since 2020.²

Fieldhand AI does the sorting, and tells you whether zones pay on your field.

¹ Industry figures, not Aiterra field results. Source: USDA ERS, McFadden, Njuki and Griffin, EIB-248 (2023). ² CropLife/Purdue Precision Dealer Survey, 2025: 43% of acres, against 44% in 2020.

How it works

Ask. See. Act.

Field soil reports on demand. Seed, fertilizer and lime prescriptions. What-if questions, recalculated with the same tools. Nothing is ordered without the grower’s yes.

01
Ask
Ask about a soil test, a field, or the season ahead. Then ask what if: a different yield goal, manure credit or lime product. The tools recalculate the rate.
fieldhand.aiterra.ag
I farm in Iowa. My latest soil report gives phosphorus at 12.5 ppm by Bray-1. I'm growing corn aiming for 210 bu/ac. Where does that put me, and what should I do about P this season?
At 12.5 ppm Bray-1, your phosphorus is in the Low category based on Iowa State University guidelines (optimum is 16–20 ppm). To support a 210 bu/ac corn crop and start building your soil test levels, plan a build-and-maintain application this fall.
Target rate: 100 to 115 lb P₂O₅ per acre (covers the ~67 lb your corn will remove, plus build-up).
Product equivalent: roughly 200–220 lb/ac of MAP (11-52-0) or 220–250 lb/ac of DAP (18-46-0).

SourcesIowa State PM 1688
What if I credit 4,000 gal/ac of hog manure? The analysis shows 25 lb P₂O₅ per 1,000 gal.

Recalculated with the same tools: 4,000 gal × 25 lb per 1,000 gal is a 100 lb P₂O₅/ac credit. That covers most of your 100 to 115 lb target, so commercial P drops to 0–15 lb/ac.

Recalculated by toolExample follow-up. The manure analysis is an example input.

Order a soil testOrder sampling for this field. The request goes to the lab by email.Generate a prescriptionSeed, fertilizer or lime, zone by zone. For this field: 100–115 lb P₂O₅/ac.Share it with an expertEmail the prescription to your applicator, agronomist, crop specialist or soil scientist. They can call you and come out to the field.
02
See
The answer becomes a seed, fertilizer or lime prescription by zone, or a single rate when zones don’t pay. Every number is there to check.
fieldhand.aiterra.ag
Phosphorus prescriptionExample field · 80 ac · fall application
Awaiting approval
ZoneArealb P₂O₅/ac
Zone 122 ac125
Zone 230 ac110
Zone 318 ac100
Zone 410 ac70

Field average
107 lb/ac
Total
8,550 lb P₂O₅
Edit ratesApprove
03
Act
Approve it, and the signed record goes by email to whoever does the work. It stays with the field for next season’s question.

fieldhand.aiterra.ag

Prescription recordExample field · 80 ac
Signed

Plan
P build and maintain, 4 zones

Rates
70 to 125 lb P₂O₅/ac by zone

Source
Iowa State PM 1688

Approved
By the grower, in the app

Signed · Reviewed by R. MartinThe record stays with the field.

Shared by email invite with the applicator

From a soil test to work in the field, with the grower’s yes at each step.

Download · PDF, 50 KB

Aiterra in Brief

The problem, the product and how we test it, in two pages.

Download the data sheet (PDF) →

Explore the product →

Fieldhand EarnedZones · Field A, Northern Iowa

Zones where they pay. One rate where they don’t.

Field A is a real Northern Iowa field, names removed. It was grid-sampled at the same 35 grid points in March 2024, October 2024 and April 2025. In April 2025, pH ran 5.6–7.3, K 98–278 mg/kg, P 17–69 mg/kg and organic matter 1.0–5.5%. Fieldhand’s call: zones for pH, potassium and organic matter, and one rate for phosphorus.

18 of 18held-out tests where a smooth map had positive skill over the field average
5 of 18where a two-zone map cleared the 95% confidence bar
0for phosphorus, so one rate was the honest answer

Issued US patents

Read the science

Download the Field A trial report (PDF, names removed, map mirrored)

pHFieldhand’s call: Zones

Mar 20245.3–7.0median 5.9Not clear
Oct 20245.5–6.9median 6.1Not clear
Apr 20255.6–7.3median 6.1Zones pay

In April 2025, a two-zone map beat the field average on held-out samples, and the pattern held between dates. The five lowest April values, pH 5.6 to 5.7, sit at one end of the field.

Spring 2024 to spring 2025: pH rose at 28 of 35 points (median +0.2).

Potassium mg/kgFieldhand’s call: Zones

Mar 202480–375median 155Not clear
Oct 202485–338median 160Not clear
Apr 202598–278median 169Zones pay

The April 2025 test favored two zones over one rate, and the pattern held between dates.

Over the year, K rose at 25 of 35 points (median +21 mg/kg).

Phosphorus mg/kg, Bray P1Fieldhand’s call: One rate

Mar 202411–85median 26Not clear
Oct 202412–87median 35Not clear
Apr 202517–69median 35Not clear

No zone map cleared the bar on any date, and the pattern shifted between dates. Fieldhand recommends a single rate until more samples show a pattern.

Over the same year, P rose at 22 of 35 points (median +4 mg/kg).

Range across the 35 points, with the medianDid two zones beat one rate on held-out samples?Map: Schematic, not the field’s boundary

Science you can check

The model talks. The math is computed.

The agent handles the conversation. Rates and conversions are computed by software tools from published soil-science equations and regional agronomic guidelines (in the US, land-grant university tables), not generated by the language model, so the same inputs give the same numbers and each number can be traced.

01You askA question about your field, in plain words.
02The agentHandles the conversation and calls the tools.
03The toolsApply published equations and regional agronomic guidelines to compute rates and conversions.
04The answerSame inputs, same numbers, each traced to its source.

93.4%

Ranges that hold

Stated 90% ranges held 93.4% of measured lab values across 743 Upper Midwest points, and 94.2% across 2,284 Pacific Northwest points. On Field A, 197 of 210 later lab values fell inside their ranges.²

One rate

When zones don’t pay

Before recommending zones, Fieldhand checks on held-out samples whether they beat a single field-average rate. If they don’t, it recommends a single rate and says why. See Field A

77,639

Real lab values

Lab values across 41 analytes, from about 13,630 acres and 137 fields in three US regions.²

AgBench

Hard questions, logged with their tools.

AgBench is our evaluation of hard agronomic questions, posed as hypothetical field scenarios. Each answer is logged with the tools Fieldhand called. Full results publish when the evaluation is complete.

² Next proof point: we are working with our growers to bring in yield, fertilizer-use and lime-use records, and we are adding those outcomes to the validation as each season’s data arrives. These figures are Aiterra’s own; we welcome independent review.

Who it’s for