Personalized care instructions work better when they account for someone's actual habits, literacy level, and daily routine, and AI can adapt better to this when you give it examples of instructions that landed well before. Showing what resonated in past guidance helps the AI tailor new instructions to real humans rather than generic patients.
Few-shot prompting is a technique where you provide an AI with two or three examples of the output style or format you want before asking it to generate new content, training it on your specific preferences without any technical setup. By showing the AI what a good care instruction, daily checklist, or handoff note looks like for your situation, you get outputs that match your real-world needs from the start.
Caregivers who need consistent, personalized documentation across multiple providers or family members benefit greatly from this approach. It eliminates the guesswork of describing what you want in the abstract and produces results that feel tailored to your loved one rather than generic.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.