Security
Your farm data stays under your control with private deployments, role-based access, and auditable feeding decisions.
Feed just right.
Every bite.
FeedRight reads live signals from the water and the fish, then feeds the right amount at the right moment — with your team in control.
Trusted by industry leaders and developers worldwide
Safe. Flexible. Independent.
Deployment
Run FeedRight in the cloud, inside a virtual private network, or connected to the camera and sensor stack already on site.
Customization
Train on your own cage history, species profiles, feed curves, and seasonal water conditions for site-specific automation.
FeedRight
Your sovereign aquaculture workplace
All your cages connected, nothing outsourced. FeedRight brings every feeding signal together on your terms.
Open dashboardAutomations
Browse workflows that streamline repetitive feeding operations with review-ready controls.
DiscoveryRunsMy buildsMonitor
+2
Waste Scope
Measure pellet drift and surface activity for the feeding team.
+2
Meeting Summarizer
Turn operator handover notes into action items.
+2
Cage Review
Review complex water-quality changes before automation.
+2
Customer Overview
Get a holistic view of farm and cage performance.
Powering progress across aquaculture
White Travelly
Appetite-aware feeding for cage, pond, and raceway teams.
Offshore cages
Open-water cage operations with video, oxygen, and feed telemetry.
Shrimp operations
Pond-level control for feed timing, aeration, and growth variance.
Hatcheries
Early-stage rearing workflows with consistent checks and audit trails.
Pond grow-out
Multi-pond feeding plans with site-level visibility for supervisors.
Circular net pens
Dense cage arrays monitored as one coordinated operating surface.
Lake farms
Freshwater cages planned around site conditions and farm routines.
Blue-water arrays
Large sea-farm layouts with consistent feed and performance tracking.
Validated today for white trevally. Other species and environments are rolling out next.
The FeedRight loop
See the cage. Decide the feed. Prove it first.
Mythos, Argus Vision, and Sim Lab aren't three tools to stitch together — they're one feeding loop. Argus Vision turns underwater video into appetite and waste signals, Mythos weighs them against oxygen, biomass, and the forecast to size each feed, and Sim Lab rehearses the strategy against modeled fish before a pellet ever drops. Every automated action still waits on your team's review.
01 · See
Argus Vision
Computer vision that reads appetite, pellet drift, and uneaten feed from the cage's own underwater cameras, in real time.
- Works with the underwater camera streams already on site
- Flags low visibility before a decision is trusted
- Turns raw video into feeding confidence scores
02 · Decide
Mythos
An AI feeding agent that weighs Argus's signals against oxygen, biomass, current, and feed history to size each meal as a ration of live biomass — then waits for your review.
- Doses a continuous ration of biomass, so kg scales with fish size
- Blocks or trims any dose past oxygen, current, and ration limits
- One shared model that keeps learning across cages and species
MYTHOS AGENTARGUS + SIM INPUTS
Plan tomorrow's feed for Cage B using camera confidence, dissolved oxygen, appetite, biomass, and forecast wind.
05:30Light feed1.2 kg
08:00Medium feed2.4 kg
12:30PauseLow oxygen watch
16:00Heavy feed3.5 kg
Argus · See→Mythos · Decide→Sim Lab · Prove
03 · Prove
Sim Lab
A high-fidelity cage simulation that rehearses a feeding strategy against modeled fish, pellets, and water before it ever reaches the water.
- Rehearses feeding strategies before field rollout
- Models cage conditions, fish response, and pellets
- Gives developers repeatable test scenarios
Why leading farms trust FeedRight
←→
Results
FeedRight gives operators a shared source of truth for appetite, safety, and feed cost without giving up control of farm data.
Measured across high-frequency feeding workflows.
One operating layer for cage video, water quality, feeding action, and audit trail.
Pricing
We bill 20% of the feed we save. You keep 80%.
No per-cage licence. Mythos prices itself off the feed cost it removes — 20% of the saving, or $150 per cage a month, whichever is greater — so above the floor your return is pure arithmetic: 1 ÷ 0.20 = 5.0×.4
The target
Feed is the cost
Feed is roughly half of a cage farm's operating cost — the one big number worth moving. Mythos is built to attack it, not to stack another monthly licence on top.
feed ÷ total farm cost
≈ 50%
The saving
~14% less feed, same harvest
Appetite-aware dosing holds economic FCR near 1.08 against a ~1.25 regional baseline — about 14% less feed for the same fish out.
production × 1.25 FCR × $1,400/t × 14%
= feed saved per year
The share
You keep 80%
We bill 20% of the feed we save — or $150 per cage a month, whichever is greater. Above that floor it's a flat 5.0× return whether you run 6 cages or 600.
1 ÷ 0.20
= 5.0× return
Worked example
500 T/yr white trevally · 12 cages · $1,400/t feed
Baseline feed spend
$875,000
at 1.25 FCR
Feed saved · 14%
$122,500
FCR 1.08 vs 1.25
Mythos fee · 20%
$24,500
floor $21,600/yr
You keep · 80%
$98,000
kept every year
5.0×
back to the farm for every $1 of fee
Vision-only setup
$9,000
12 cages × $750 · pays back ~4.8 weeks
+ Water-quality setup
$67,800
12 × $5,650 · pays back ~8.3 months
What it costs to switch a cage on
Mythos runs on one 2D camera per cage. Water-quality sensing is optional — add it only where oxygen or current risk demands the extra guardrails.
Vision-only Mythos
Most commonSwitch on the cages you already operate.
$750/ cage, one-time
Pays back in weeks
- $250 MizLinx 4K marine PoE camera per cage
- $500 PoE run, marine enclosure & commissioning
- Monocular AI depth — no extra sensors required
Mythos + water-quality sensing
For sites with oxygen or current risk.
$5,650/ cage, one-time
Pays back in months
- Everything in vision-only
- + $4,900 DO · temperature · salinity sonde per cage
- Hard O₂ / current safety guardrails on every dose
Mythos bills 20% of the feed saved or $150 per cage a month, whichever is greater. Per-cage setup figures are camera + install (+ sonde); shared site infrastructure — edge compute, gateway, router — is additional and scales slowly with cage count. ROI and payback are computed from the feed savings the farm keeps after our fee.
Operator resources
Find everything you need to run the platform, from model governance to calibration guides and integration docs.
Introduction
Creating a cage
Model overview
Safety limits
Release notes
CREATE A CLIENT
Connect farm data
const farm = await feedright.connect();
await farm.cages.watch("cage-b");
await farm.models.review("mythos");
Camera ready
Sensors synced
Safety active
“FeedRight gave our operators a shared source of truth for appetite, safety, and feed cost without giving up control of farm data.”
— Operations Director, coastal aquaculture group
The latest news
Curious how we got here?
From a spontaneous lunch to winning the Ocean Hackathon Global Finale — hear the story behind FeedRight.
- 1
70% reduction in feeding labor. Operator-reported reduction in feeding labor from AI-camera adaptive feeding at a cage farm in Malaysia — the closest analog to Mythos's vision-driven dosing — consistent with the "up to 70%" ceiling reported for on-demand acoustic/vision feeders in industry reviews. Reviewed automation replaces manual cage-by-cage feeding rounds; earlier non-AI automated feeders report 30–40%. A field-reported figure, not a Sim Lab model. ↩
- 2
21% less feed wasted per cage. Reduction in uneaten-pellet mass per feed window in Sim Lab versus a blind, schedule-based baseline, measured from simulated pellet tracking (pellets sinking past the school before ingestion). This is a waste-mass figure, reported separately from the FCR and feed-cost numbers. ↩
- 3
14% FCR improvement in year one. Modeled economic FCR of 1.08 against a ~1.25 regional baseline — 1 − 1.08 ÷ 1.25 ≈ 14% — from Sim Lab's bioenergetic intake-and-growth model, which is also the projected feed-cost reduction per kg at a fixed pellet price. An engineering target, not a measured grow-out result; the species' biological conversion floor (~1.0) caps further gains. ↩
- 4
We bill 20% of the feed we save, or $150/cage/month — whichever is greater. Annual saving = production × 1.25 baseline FCR × $1,400/tonne feed × 14% reduction. Mythos bills the greater of a 20% value share and a $150-per-cage-per-month floor, so wherever the share applies the return is 1 ÷ 0.20 = 5.0×. At the modeled savings the share clears the floor at the mid and large tiers; at the small tier the floor sets the fee (it runs just above 20% of the saving), which is the case the floor exists for. One-time setup is $750/cage vision-only ($250 MizLinx 4K marine camera + $500 install) or $5,650/cage with a DO/temp/salinity sonde; payback is measured against the savings kept after our fee — roughly 4–6 weeks vision-only and 7–11 months with sensors. Per-cage figures exclude shared site infrastructure (edge compute, gateway, router). Worked examples use 200/500/2,000 T-per-year sites; see /products/mythos for the full cost model. ↩