What the AI found
“AI found that your lowest energy days consistently followed evenings with red wine consumption, while your assumed trigger, caffeine, showed no statistically significant correlation.”
Before
Disorganised symptom notes and unproven assumptions
After
Actionable insights for targeted weekly hormone support
The same system, three states — real screens, not a screenshot
- 14 Nov: Low energy, headache. Diet: coffee x2, pasta, red wine. Sleep: 6h.
- 15 Nov: Good energy. Diet: tea x1, salad, chicken. Sleep: 7.5h. Cycle Day 18.
- 16 Nov: Exhausted. Diet: coffee x3, pizza, red wine. Sleep: 5.5h.
- 17 Nov: Moderate energy. Diet: coffee x1, fish, veg. Sleep: 7h. Cycle Day 20.
Prompt
You are an expert data analyst for health and wellness data. I have exported my daily tracking data for the past 6 weeks into a Google Sheet, with columns for Date, Cycle Day, Sleep Duration (hours), Energy (1-5, 5=best), Headache (Y/N), Diet Notes, and Alcohol (Y/N, Type). My primary concern is low energy. Analyse this data and identify the strongest correlational patterns between my logged activities (diet, sleep, alcohol, cycle day) and my reported energy levels. Quantify any significant correlations you find. Focus on non-obvious patterns.
Can you help me find patterns in my perimenopause tracking data to understand my low energy days?
AI
I can. I've analysed your data for the past 6 weeks. Your three worst-energy days (rating 1 or 2) all followed evenings with red wine consumption, occurring within 8-12 hours of drinking. Conversely, caffeine intake showed no statistically significant correlation with your reported low energy levels, contradicting your initial assumption. Sleep duration below 6 hours correlated with a 1.5 average drop in energy rating.1.8 (on scale of 5)
Avg. Energy Post-Wine
3.9 (on scale of 5)
Avg. Energy Post-Caffeine
1.5 pts
Energy Drop <6h Sleep
One Hormone Cycle AI Audit
From scattered daily notes to a targeted weekly hormone support plan, informed by AI-driven pattern recognition.
A 38-year-old marketing manager, Northern Europe, managing perimenopause symptoms.
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
Eleanor diligently tracked her perimenopause symptoms, diet, and lifestyle factors in a Notion document. Despite weeks of data entry, she felt overwhelmed by the sheer volume of information. She suspected caffeine was her biggest energy drain, but couldn't pinpoint any clear patterns or connections in her unstructured notes. The data felt inert, offering no actionable path forward.
- 14 Nov: Low energy, headache. Diet: coffee x2, pasta, red wine. Sleep: 6h.
- 15 Nov: Good energy. Diet: tea x1, salad, chicken. Sleep: 7.5h. Cycle Day 18.
- 16 Nov: Exhausted. Diet: coffee x3, pizza, red wine. Sleep: 5.5h.
- 17 Nov: Moderate energy. Diet: coffee x1, fish, veg. Sleep: 7h. Cycle Day 20.
- 18 Nov: Very low energy, brain fog. Diet: coffee x2, burger, red wine. Sleep: 6h.
Working state
Resources, doing its job
Using a custom Wellness & AI playbook, Eleanor exported her Notion data into a Google Sheet. She then leveraged ChatGPT to analyse the anonymised, structured data. The prompt clearly outlined the task: identify correlations between her logged activities and symptom severity, specifically energy levels. The AI processed the data, revealing a surpising, counter-intuitive insight.
Prompt
You are an expert data analyst for health and wellness data. I have exported my daily tracking data for the past 6 weeks into a Google Sheet, with columns for Date, Cycle Day, Sleep Duration (hours), Energy (1-5, 5=best), Headache (Y/N), Diet Notes, and Alcohol (Y/N, Type). My primary concern is low energy. Analyse this data and identify the strongest correlational patterns between my logged activities (diet, sleep, alcohol, cycle day) and my reported energy levels. Quantify any significant correlations you find. Focus on non-obvious patterns.
Can you help me find patterns in my perimenopause tracking data to understand my low energy days?
AI
I can. I've analysed your data for the past 6 weeks. Your three worst-energy days (rating 1 or 2) all followed evenings with red wine consumption, occurring within 8-12 hours of drinking. Conversely, caffeine intake showed no statistically significant correlation with your reported low energy levels, contradicting your initial assumption. Sleep duration below 6 hours correlated with a 1.5 average drop in energy rating.Use case implemented
The finished system, running on its own
With the AI's insight, Eleanor adjusted her evening routine, prioritising alternatives to red wine. She continued using her Notion tracker, but now exported the data weekly into the Google Sheet for a quick AI audit. This streamlined process provided a clear, evidence-based feedback loop, transforming passive tracking into proactive, targeted self-support, yielding better energy days.
1.8 (on scale of 5)
Avg. Energy Post-Wine
3.9 (on scale of 5)
Avg. Energy Post-Caffeine
1.5 pts
Energy Drop <6h Sleep
What an outside observer would notice
From ~60 min to 10 min
Time spent analysing symptoms per week
Reduced by 35%
Number of sustained low-energy days per month
Dramatic improvement
Clarity on symptom triggers
The stack — build it yourself
Flexible, customisable for diverse symptom and lifestyle logging.
Universal for structuring and cleaning exported data for AI analysis.
Excellent for conversational analysis of structured data, identifying correlations.
These are the tools used in this story. Any can be swapped for an equivalent you already trust.
Go deeper
Do this yourself
See Resources
This story runs on Resources. The tools and prompts above are the real build — swap any tool for your own equivalent and follow the same steps.