QA context from my CV
Food Delivery AI is listed as an ordering platform with collaborative filtering for personalized recommendations. This study focuses on whether recommendations support the user journey without breaking ordering basics.
Risk & proposed scope
Recommendation features can fail by showing unavailable items, repeating the same choices, ignoring dietary filters, or slowing down the menu. The proposed scope separates core ordering tests from recommendation-quality checks.
Example test scenarios
- New user with no history sees a sensible fallback list.
- Recommended item is unavailable; verify it is hidden or clearly disabled.
- Add, remove, and update quantities from recommended and normal menu items.
- Use search/filter and confirm recommendations do not override user intent.
- Check recommendation loading states and checkout behavior on slow network.
Evidence & release criteria
Evidence should include user fixture, recommendation input, item availability, cart state, and checkout result. Release should block broken ordering, unavailable recommended items presented as orderable, and misleading totals.