Search Feature Implementation: Database vs Search Engine
Choose and implement search appropriately.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
ROLE: You are a backend architect. CONTEXT: Content or data to search: [CONTENT_TYPE] with [RECORD_COUNT] records. Query types (keywords, filters, typo tolerance, ranking, autocomplete): [QUERY_TYPES]. Stack and database: [STACK]. Team capacity to run another service: [OPERATIONS_CAPACITY]. Update frequency: [UPDATE_FREQUENCY]. STEPS: 1. Compare database full-text search, a dedicated search engine, and a hosted search service on relevance, features, cost, operations, and consistency. 2. Recommend one for my case and explain what would change the recommendation. 3. Design the indexing pipeline, the fields and weights, synonyms, and handling of deletions. 4. Provide a query example and a ranking strategy. 5. Plan the interface: search box, suggestions, results, empty state, and filters. 6. Define metrics: zero-result rate, click position, and search exits. OUTPUT FORMAT: Comparison table, recommendation, indexing design, query example, UI plan, and metrics. QUALITY BAR: Start simple. Measure before adding complexity. Plan for privacy in search logs. FOLLOW-UP: Ask me what people search for most.
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.