What the AI found
“AI found a consistent 15% drop in client morning glucose readings on days following a specific meal composition you noted simply as "balanced plate" – a pattern you'd missed manually scanning weeks of client logs.”
Before
Ad-hoc client notes, no structured review
After
Structured weekly metabolic insights, 15 min.
The same system, three states — real screens, not a screenshot
- Client: J.M. (Metabolic)
- 2024-03-04: Morning glucose 6.1 mmol/L. Meals: Oats, berries. Lunch: Chicken salad. Dinner: Lentils, veg. Felt good.
- 2024-03-05: Morning glucose 5.8 mmol/L. Meals: Eggs, avocado. Lunch: Balanced plate. Dinner: Fish, sweet potato. Good energy.
- 2024-03-06: Morning glucose 6.3 mmol/L. Meals: Cereal, banana. Lunch: Sandwich. Dinner: Pasta, sauce. Bit sluggish P.M.
Prompt
Analyse the provided client notes. Identify any recurring dietary patterns that consistently precede a morning glucose reading change of +/- 0.5 mmol/L. Only report quantitative findings.
Here are J.M.'s notes for the week: 2024-03-04: Morning glucose 6.1 mmol/L. Meals: Oats, berries. Lunch: Chicken salad. Dinner: Lentils, veg. Felt good. 2024-03-05: Morning glucose 5.8 mmol/L. Meals: Eggs, avocado. Lunch: Balanced plate. Dinner: Fish, sweet potato. Good energy. 2024-03-06: Morning glucose 6.3 mmol/L. Meals: Cereal, banana. Lunch: Sandwich. Dinner: Pasta, sauce. Bit sluggish P.M. 2024-03-07: Morning glucose 5.0 mmol/L. Meals: Eggs, avocado. Lunch: Balanced plate. Dinner: Tofu stir-fry. Excellent. Please analyse.
| Client ID | JM001 |
| Week of | March 4, 2024 |
| Morning Glucose Avg | 5.8 mmol/L |
Weekly Metabolic Review in Apple Notes with AI
A nutritionist moves from scattered daily notes to a structured weekly metabolic review, powered by AI within Apple Notes.
A nutritionist running a small EU practice, focusing on metabolic health.
Tools used
The real tools used here — swap any for your own equivalent. Each links to how we’d set it up.
Starting state
Before anything was set up
For weeks, client progress existed as unstructured daily notes within Apple Notes. While thorough, the sheer volume of text made it challenging to spot emerging patterns across multiple clients. Trends remained hidden, buried in individual entries, requiring time-consuming manual review that rarely happened.
- Client: J.M. (Metabolic)
- 2024-03-04: Morning glucose 6.1 mmol/L. Meals: Oats, berries. Lunch: Chicken salad. Dinner: Lentils, veg. Felt good.
- 2024-03-05: Morning glucose 5.8 mmol/L. Meals: Eggs, avocado. Lunch: Balanced plate. Dinner: Fish, sweet potato. Good energy.
- 2024-03-06: Morning glucose 6.3 mmol/L. Meals: Cereal, banana. Lunch: Sandwich. Dinner: Pasta, sauce. Bit sluggish P.M.
- 2024-03-07: Morning glucose 5.0 mmol/L. Meals: Eggs, avocado. Lunch: Balanced plate. Dinner: Tofu stir-fry. Excellent.
Working state
Membership, doing its job
To surface these hidden trends, the nutritionist copied a week's worth of a client's dietary and biometric data from Apple Notes and pasted it directly into a chat window. Alongside this data, they provided a specific prompt, asking the AI to cross-reference dietary notes with biometric readings, looking for metabolic patterns.
Prompt
Analyse the provided client notes. Identify any recurring dietary patterns that consistently precede a morning glucose reading change of +/- 0.5 mmol/L. Only report quantitative findings.
Here are J.M.'s notes for the week: 2024-03-04: Morning glucose 6.1 mmol/L. Meals: Oats, berries. Lunch: Chicken salad. Dinner: Lentils, veg. Felt good. 2024-03-05: Morning glucose 5.8 mmol/L. Meals: Eggs, avocado. Lunch: Balanced plate. Dinner: Fish, sweet potato. Good energy. 2024-03-06: Morning glucose 6.3 mmol/L. Meals: Cereal, banana. Lunch: Sandwich. Dinner: Pasta, sauce. Bit sluggish P.M. 2024-03-07: Morning glucose 5.0 mmol/L. Meals: Eggs, avocado. Lunch: Balanced plate. Dinner: Tofu stir-fry. Excellent. Please analyse.
Use case implemented
The finished system, running on its own
The system now runs efficiently each Sunday morning. The nutritionist dedicates 15 minutes to copying the previous week's notes for each client into the AI. The AI’s summarised metabolic insights are then appended to the client's file, forming a reliable, structured weekly review that informs subsequent client consultations.
| Client ID | JM001 |
| Week of | March 4, 2024 |
| Morning Glucose Avg | 5.8 mmol/L |
| Identified Pattern | Post 'balanced plate' lunch, average morning glucose consistently decreased by 15.3% (from 6.1/6.3 to 5.0/5.8 mmol/L). |
| Actionable Insight | Explore specific macronutrient breakdown of 'balanced plate' for replication guidance. |
What an outside observer would notice
45 minutes/week
Time saved per client review
Increased
Pattern detection accuracy
1-2 per review
Client metabolic insights
The stack — build it yourself
Ubiquitous, quick for on-the-go notes, already part of existing workflow.
Provides nuanced pattern recognition across qualitative and quantitative data.
Ideal for quantitative tracking and easy comparison over time.
These are the tools used in this story. Any can be swapped for an equivalent you already trust.
Go deeper
Do this yourself
See Membership
This story runs on Membership. The tools and prompts above are the real build — swap any tool for your own equivalent and follow the same steps.