Autrix reads sonar returns, vessel tracks, landings and satellite feeds together — and hands fishery managers a stock assessment, an enforcement flag and a quota recommendation they can sign.
Landings in Sector 7 are down 18% this quarter and three trawlers went dark near the marine reserve boundary. Tell me what's happening, and what I should authorise for Q4.
Autrix sits on top of the data your agency already collects — and the satellite data it doesn't. Three stages, one audit trail, no research team required to operate it.
Read the technical brief
Vessel tracks, echosounder returns, port landings, observer reports, SAR passes and satellite temperature, normalised into one vessel-and-stock model. Silence is treated as data, not a missing row.
Surplus-production and length-based assessments run weekly instead of every third year, with uncertainty carried all the way through to the number a manager actually acts on.
Every advisory exports with its inputs, model version and confidence band intact — so it survives a review board, a court hearing or a regional commission meeting without being re-litigated.
One data spine, deployed per nation, per regional commission or per farm. No pilot that dies in a research notebook — the outputs land in the workflow your team already runs.
Talk to the founder
Biomass, recruitment and reference points per stock, refreshed as the data lands rather than when the survey vessel returns. Trends are visible while there is still a season left to change.
Cage 4 growth is flat for nine days. Feed or water?
Water. Dissolved oxygen dipped below 4.1 mg/L on six nights; feed conversion held. Cut the night ration 15% and aerate from 02:00.
Managers, farm operators and enforcement officers query the same model in words, not SQL — and every answer comes back with the records and the run behind it, never an unsourced claim.
Continuous monitoring scores loitering, transshipment and reserve incursion the moment a signal drops, then hands enforcement a ranked tasking list instead of a raw position dump.
Biomass, recruitment and reference points from sonar, catch and effort data — weekly, with uncertainty intact rather than averaged away.
AIS gap analysis fused with SAR and VMS to score loitering, transshipment and reserve incursion before the catch reaches a port.
Ration schedules, growth curves and mortality risk tuned per cage from biomass sensors, water quality and local weather.
Temperature, chlorophyll, salinity and upwelling anomalies projected 90 days out, mapped onto the grounds your fleet actually works.
Krish started Autrix from a pattern that repeats on almost every coastline: the signals needed to prevent a collapse already exist — scattered across port ledgers, vessel transponders, echosounders and satellite passes — and no agency has the tooling to assemble them before the season closes.
Autrix is the answer to that gap. Not a research model that ends as a paper, but a deployable system that fishery departments, enforcement teams and aquaculture operators can point at their own waters and use from the first week.
"Every stock collapse on record left a data trail. The collapse wasn't inevitable — the blindness was."
Fleet positions, landings records and enforcement referrals are sensitive national data. Autrix deploys inside your environment or a region you nominate, with per-tenant isolation, role-based access and an immutable log of every model run, query and export. Customer data is never used to train shared models.
