A passive, continuous labor-and-harvest intelligence layer for any operation that picks by hand, capturing what happens between the weigh and scan events existing systems already record.
Muse Ag sees everything in between.
The Gap
Existing tools such as PickTrace and FairPick capture weighing and scanning: the moment a harvest bin or unit is recorded. They do not see the continuous movement in between, including dwell, route, walking time, and non-productive minutes.
That between-events signal is where harvest cost quickly accumulates, and no incumbent is built to read it and leverage the potential from it.
The Wedge
Muse Ag runs a passive positioning layer across the picking environment, resolving worker movement to the row, bay, or zone, with no change to how crews already work.
One unified system across outdoor field, polytunnel, and greenhouse: same hardware, same schema, same interface. Only the calibration differs. No competitor spans all three by design.
One Capture · Four Lenses
The farmer is why this gets adopted. The same passive capture happens to serve three others at no added effort.
The Moat
The defensible asset is the data spine, not the reasoning layer. Clean per-worker, per-block capture compounds in value with every season, a network effect competitors cannot retrofit.
Privacy By Design
Dual identity: the employer side sees anonymous worker numbers only. Workers who opt in claim and own their record, sharing identity on their own terms. A cryptographically assured guarantee, not a policy promise.
Operations intelligence, not individual surveillance. Aggregate and block-level by default.
Muse Ag
Field Survey
36.97°N 121.90°W
North Star · 07E
Active picking ratio · today
71%
24 pickers on field · 6 blocks surveyed · data quality 96%
3,108
Picker-rows
2W
Flagged
+2.4%
Vs last shift
Movement intelligence
passive · always on
Block 7 East
18 rows · 9 pickers today
Rows by yield rate · tap a row
Picker #1207 · movement this shift (illustrative)
Pace
strong
Long dwell
row end
A long dwell at one row-end, repeated across different pickers, points at the row, a drainage or access issue, not a slow worker. The pattern points at the place, not the person.
Plant issue
Block 7 East · row 11 end · 9:14a
Possible early leaf curl. Suggest agronomist review. Suggested, not diagnosed.
Flagged by picker #1188 · anonymous
Your pickers already look at every plant in every row for hours a day. The flag turns the best eyes on the farm into a scouting network: one tap, located automatically, no extra walk. Passive movement is the spine; this is the signal only human eyes catch.
Export · audit-ready
Traceability record
Generated from the same shift data. No personal worker information leaves the operation: anonymous tokens only.
The buyer gets traceability. The contractor gets compliance. Same capture, different lens.
The numerator behind every kg/hr
How harvest data comes in
Movement tells Muse Ag where each anonymous worker was and for how long. Harvest weight tells it how much came off. Marry the two and you get yield per row, per worker. Muse Ag takes the weight in however your operation already records it, with no new step at the weigh station.
Three ways in, pick what fits the operation
Existing system feed · preferred
If the operation already runs PickTrace, FairPick, or similar, Muse Ag ingests the weigh events directly. Nothing changes for the crew. This is the integrate path.
Scan a tally sheet
Where weights are still on paper, the checker photographs the tally and Muse Ag reads it. The same sheet they already fill out, nothing new to learn.
Manual entry
Small operations can key weights in by block or lot. The fallback that always works.
Weight meets movement
Harvest weight
how much came off
Movement
where & how long
Yield per row, per worker
the view the dashboard is built on
Agronomist view · Block 2 West
A signal worth a look
Muse Ag does not diagnose. It gets you to the right spot earlier, with the crew's eyes and the field context already assembled. The call is still yours.
WHAT IS KNOWN · from the field
Corroborated crew flag
3 different pickers flagged pests in the southeast corner, rows 12 to 14, across two days. Independent eyes, same spot. This is signal, not anecdote.
Movement context
Dwell in that corner has crept up over the same five shifts the flags appeared, and Block 2 West yield rate is trending down. The movement layer and the crew flags point at the same place.
Current conditions · observed, illustrative
Temp
74°F
Humidity
82%
Wind
6 mph
UV / DLI
6 · 38
Warm, humid, low wind, three days running. You know what that combination tends to favor. The conditions are context you read, not a verdict Muse makes.
Recent scouting history
No flags in this corner the prior two weeks. The cluster is new, which is exactly when early action is cheapest.
WHAT IS MODELED · forward-looking
external agent models · off by default
Predictive outlook · hypothesis
An external model reading these conditions plus the flag pattern projects elevated spread risk over the next 5 to 7 days if untreated. A forecast to weigh, not a fact. You confirm it on the ground.
Photo triage flag · elevated
An initial scan of #1207's photo shows visual markers worth a close expert look. A prompt to prioritize, never a named diagnosis. The expert names it; the scan only raises a hand.
The data is the moat; the prediction rides on top of it. The more low-touch field data Muse Ag captures over time, the better any agent model performs. These outputs are illustrative of that direction, not a capability claim today.
Your call
Whatever you decide, Muse Ag logs it against the flag and closes the loop, so the crew sees their flag mattered and the record shows the signal-to-decision time.
Why the farmer adopts
Where the savings start
The clearest early return is not squeezing pickers. It is turning the crew into a sensor network, so a field problem surfaces early and gets treated precisely. The first dollars are the agronomist's scarce time and the cost of a late, broad intervention instead of an early, narrow one.
A problem spreads with each day it goes unseen. Crew flags it early; the scheduled scout walk finds it late.
Early catch · treated area
0.1 ac
Late catch · treated area
0.9 ac
Strawberry gross value per acre varies widely. Set this to the operation's real figure.
Illustrative value protected, per incident
$19k
= (late area − early area) × value/acre × assumed loss in the affected zone. Loss rate and spread are modeled, not measured.
Without Muse Ag
Walks every block on a schedule, hunting for what might be wrong. Scarce expert time spent searching.
With Muse Ag
Directed straight to where the crew flagged a corroborated signal. Time goes to diagnosis and decision, not the hunt. Earlier eyes, earlier call.
2026 · strawberries
North Star Farms
Amount picked
31 t
Avg picking rate
8.9 kg/hr
Rate through the season (illustrative)
This is María's own record: her proof of skill, carried season to season. The employer never holds the named version; she decides who sees it.
Tap a block on the map to drill into rows and an anonymous picker's movement path. Switch the date range to see the data compound. All figures are illustrative placeholders.
Demo settings
Presenter tool, not part of the product. Adjust the core inputs and the headline figures move with them. Everything stays illustrative.