Design the data flow from USDA nutrition data into a Google nutrition experience

You are a product manager at Google. Explain how you would design the algorithm that takes nutrition data published by the USDA (the US Department of Agriculture) and displays it in Google's nutrition feature, covering everything from obtaining the data to what the user finally sees. ```hint Trace one food Pick a single food and follow it from the USDA's published record to the screen, noting what has to happen at each hand-off and what could go wrong there. ``` ### Clarifying Questions - Which surface does "Google nutrition" refer to: the nutrition facts shown when someone searches for a food, or a nutrition feature inside a health or fitness app? - Which USDA data is in scope: nutrient values for individual foods, dietary guidance such as recommended daily amounts by age and sex, or both? - Is this a one-time import, or must what users see stay in step with the USDA's updates? - How deep should the answer go into the system design, compared with the product decisions about what to show? ### What a Strong Answer Covers - A chosen interpretation of the surface and the USDA data, with the user need it serves - The end-to-end flow from acquiring and cleaning the data to linking it with what users ask for and presenting it - Handling of ambiguous food names, portions, and missing or conflicting values - Safeguards for accuracy, freshness and attribution suited to health information - How quality is evaluated before launch and how success is measured after it ### Follow-up Questions - How would you handle foods the USDA data does not cover well, such as regional dishes or restaurant meals? - A user reports that the calories shown for a common food are wrong. How do you investigate, and how do you stop the same kind of error from recurring? - Would you show one default serving or let users change the portion, and how would you test that choice?