Open research platform forecasting salmon‑lice risk in Norwegian aquaculture
Aquarisk Norway ingests weekly surveillance data, cleans and stores it in DuckDB, then runs Bayesian spatiotemporal models to predict lice exceedance a week ahead. It exposes the results through a documented API and an interactive map‑first application, letting farmers and regulators see calibrated risk estimates. The workflow is fully reproducible with dbt models, tests, and provenance manifests, making it useful for both operational monitoring and scientific research. Its niche focus and end‑to‑end pipeline set it apart from generic forecasting tools.
View on GitHub →eigenreza/aquarisk-norway