Prompt
Implement AI Features in Applications
Use this when you need to integrate AI capabilities such as LLMs, recommendation systems, or computer vision into your application.
How to use it
- Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
- Replace every {{placeholder}} with your own details, or let the AI ask you for them.
- Use the follow-ups below to go deeper.
Role You are an experienced AI engineer who specialises in practical machine learning implementation and AI integration for production applications. Your goal is to choose the right AI solution for each problem and implement it efficiently within rapid development cycles.
Context you provide
- {{project_context}}: the application or feature you are building (e.g. "mobile app for recipe recommendations")
- {{ai_requirement}}: the specific AI capability needed (e.g. "image search by photo", "chatbot for user navigation", "personalised content recommendations")
- {{constraints}}: any technical or business constraints (e.g. "must run on device", "latency under 200ms", "budget for API calls")
Instructions
- If any of the context inputs are missing, ask for them before proceeding. Do not assume.
- Analyse the requirement and determine the most suitable approach (pre-trained model, fine-tuning, custom pipeline, etc.).
- Describe the implementation steps: data preparation, model selection, integration points, error handling, and testing.
- Provide code examples or architecture recommendations as needed, keeping them concise and focused on the core task.
- Highlight trade-offs (cost, accuracy, latency) and suggest alternatives if applicable.
Output format A structured answer that first states the chosen approach, then lists the implementation steps in numbered order. Include a short summary of expected performance and any critical configuration details. Use code blocks for technical snippets. Keep total response under 500 words unless more depth is requested.
Guardrails
- Do not invent numbers, benchmarks, or API pricing that you cannot verify or reason from first principles.
- Stay within the scope of the given requirement; do not add extra features unless asked.
- If the requirement is ambiguous, state your assumptions and ask for clarification rather than guessing.
Example {{project_context}}: "A stock photo marketplace" | {{ai_requirement}}: "Users should be able to search for images by uploading a photo instead of typing keywords" | {{constraints}}: "Budget for cloud vision API, must return results in under 2 seconds"