Overview

Mişko manages behavioral tests run on laboratory mice from a single place. Subjects, environments, scenarios and tests are all defined and tracked through the application.

Who it is for

Biology and neuroscience laboratories that run behavioral paradigms on rodents and need a reliable record of what was tested, on which animal, with which setup, and what the result was — without managing raw video by hand.

Core concepts

• Paradigm — a scientific test type (Morris Water Maze, Open Field, Elevated Plus Maze, Rotarod). Defined in code; the catalog is fixed and stable.
• Environment — a named, concrete instance of a paradigm: the physical setup (geometry in cm, surface color/material for vision contrast, zones).
• Subject — the mouse, kept simple (code, sex, group, birth date, notes).
• Scenario — the central object. It bundles one or more environments (whose paradigms fix the metrics), so an experiment is defined once.
• Test — a subject measured against a scenario, run environment by environment (PENDING → RUNNING → DONE / FAILED). Results are data, not a pass/fail verdict.

Internal SaaS model

Mişko is deployed on-prem, one laboratory per installation. There is no public sign-up: a superadmin is bootstrapped at startup (the system generates a strong password and prints it once), and all other users are created from inside by an administrator. Access is governed by role-based permissions tied to each person’s role in the lab.

The system boundary

Mişko deliberately does not do computer vision. It is the system of record and keeps only summary metrics + artifact URLs. A separate CV service (Python / FastAPI / YOLOv8 / ByteTrack, with its own PostgreSQL) owns the heavy data and pushes a result summary back when a test ends. See Integration.