Team project · Flutter + machine learningIn progress

MeatFresh AI

A Flutter app that classifies meat freshness from photos using a machine-learning model. QA case study in progress.

Status: In progress. This page describes the plan and scope. Results and artifacts will be added only after the work is done.

Context

MeatFresh AI helps users check meat freshness by taking or selecting a photo. A Flutter app sends the image to a machine-learning model and shows a freshness classification.

Role

The QA work for this project is still being documented. This page will describe only the testing that was actually performed.

Objective

Identify the main quality risks for users: unclear or wrong results, poor handling of bad images or network errors, and integration failures between the app, API, and model.

Scope

IN SCOPE

Functional tests for image selection, capture, and classification
Validation of result states and error handling
Integration checks between the Flutter app, API, and ML inference
Regression checks after fixes

OUT OF SCOPE

Model accuracy evaluation (kept separate from UI testing; reported only with measured data)
Performance and load testing

Approach

Derive test scenarios from the app's user flows
Design test cases with equivalence partitioning and boundary values (e.g. image size and format)
Prepare sanitized test images covering expected and invalid inputs
Record defects with reproduction steps, expected vs. actual results, and severity

Artifacts

Test plan — not yet published
Selected test cases — not yet published
Defect reports — not yet published

Results

Not yet available. Results will be added only after test execution, using real counts and observations.

Reflection & next steps

Next: finish the test plan, run the first functional pass, and publish sanitized artifacts.

Back to projects Let's talk about QA opportunities