What the AI found
“"Your 3pm energy drops didn't follow carbohydrate lunches; they appeared only on days when morning client notes exceeded 45 minutes of continuous screen time."”
Before
four coffees daily, afternoon slump blamed on carbs
After
one 8-minute audit identifying the real afternoon drag
The same system, three states — real screens, not a screenshot
- Mon: 3pm brain fog bad, skipped rice, had 3rd espresso
- Wed: 2:45pm slump again. Maybe cortisol? Try keto lunch
- Thu: 3:30pm barely focused during client review
- Fri: Completely drained by 4pm, 4 espressos total
Prompt
I have pasted 21 days of daily logs tracking lunch composition, continuous morning admin minutes, total client hours, and my 3pm energy score (1-5). Cross-reference the lowest energy scores (1-2) against the other columns. What correlates most consistently?
I have pasted 21 days of daily logs tracking lunch composition, continuous morning admin minutes, total client hours, and my 3pm energy score (1-5). Cross-reference the lowest energy scores (1-2) against the other columns. What correlates most consistently?
AI
Your 3pm energy drops didn't follow carbohydrate lunches; they appeared only on days when morning client notes exceeded 45 minutes of continuous screen time. On low-carb days with >50 min admin, energy averaged 1.8/5. On high-carb days with <30 min admin, energy averaged 4.2/5.10-min admin buffer
Protocol test
1.8 / 5 (down from 4.1)
2pm-4pm fatigue score
2 daily (steady)
Espressos logged
Mapping Practice Fatigue Using the Three Layer Method
How a private practitioner audited her daily fatigue variables using literature synthesis, a personal ledger, and a single-variable protocol.
A private practice clinical nutritionist, Northern Europe
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
Her clinic diary was meticulous for clients, but her own energy log was a chaotic string of quick mobile notes. She blamed postprandial glucose dips for her persistent 3pm exhaustion, systematically eliminating carbohydrates from her midday meals. She brewed a third and fourth espresso between appointments, feeling depleted just as her final clients arrived. With variables unisolated and hypotheses untested, she treated the fatigue as an inevitable physical decline rather than an observable systemic workflow pattern.
- Mon: 3pm brain fog bad, skipped rice, had 3rd espresso
- Wed: 2:45pm slump again. Maybe cortisol? Try keto lunch
- Thu: 3:30pm barely focused during client review
- Fri: Completely drained by 4pm, 4 espressos total
Working state
Core Course, doing its job
Applying the Core Course's framework, she created a minimal three-column ledger tracking lunch composition, uninterrupted documentation minutes, and mid-afternoon focus. After three weeks of data collection, she ran the cross-variable audit prompt through Claude. The language model parsed the table, isolating the true correlate: mental fatigue from dense, uninterrupted screen charting before 1pm, completely detached from whether she ate grain bowls or salads. Her underlying assumption had been disproven by her own structured data.
Prompt
I have pasted 21 days of daily logs tracking lunch composition, continuous morning admin minutes, total client hours, and my 3pm energy score (1-5). Cross-reference the lowest energy scores (1-2) against the other columns. What correlates most consistently?
I have pasted 21 days of daily logs tracking lunch composition, continuous morning admin minutes, total client hours, and my 3pm energy score (1-5). Cross-reference the lowest energy scores (1-2) against the other columns. What correlates most consistently?
AI
Your 3pm energy drops didn't follow carbohydrate lunches; they appeared only on days when morning client notes exceeded 45 minutes of continuous screen time. On low-carb days with >50 min admin, energy averaged 1.8/5. On high-carb days with <30 min admin, energy averaged 4.2/5.Use case implemented
The finished system, running on its own
The practitioner now runs an eight-minute review every Friday afternoon. The Ledger logs client density, admin blocks, and self-reported 3pm alertness on a simple five-point scale. By treating administrative overload rather than dietary carbohydrates as the primary fatigue lever, she scheduled mandatory ten-minute walking buffers between consult blocks. Afternoon fog stopped accumulating, without altering lunch choices. The three-layer method gave her a clinical-grade observational feedback loop for her own practice day, free of guesswork.
10-min admin buffer
Protocol test
1.8 / 5 (down from 4.1)
2pm-4pm fatigue score
2 daily (steady)
Espressos logged
What an outside observer would notice
-56% across 3 weeks
Afternoon alertness drop
Halved from 4 to 2
Post-lunch espresso
The stack — build it yourself
Frictionless daily entry of consults, documentation minutes, and afternoon alertness scores.
Strong contextual reasoning over tabular data without hallucinating statistical correlations.
These are the tools used in this story. Any can be swapped for an equivalent you already trust.
Go deeper
Do this yourself
Explore the Core Course
This story runs on Core Course. The tools and prompts above are the real build — swap any tool for your own equivalent and follow the same steps.