new file mode 100644
index 0000000..785947a
@@ -0,0 +1,30 @@
+---
+type: idea
+title: personalized medicine / wellness testing
+tags: [health, science, data, software]
+status: raw
+first_captured: 2026-02-09
+updated: 2026-04-10
+sources:
+ - sources/ideaflow/2026-02-09_idea-of-things-medicine-procedures-strategies-impact-pe.md
+ - sources/ideaflow/2026-02-10_two-ways-to-approach-personalized-medicine-linked-note.md
+ - sources/google-sheets-ideas.md
+---
+
+# personalized medicine / wellness testing
+
+the core insight: "things (medicine, procedures, strategies) impact people differently" — biologically and psychologically. a unified testing framework for personalized wellness. example: go onto the platform, start testing sleep factors, and in 1 week have strong results for what most impacts your sleep out of commonly cited factors.
+
+two approaches identified: (1) ultra data collection — long-standing correlations taken at face value, run data analysis, suggest experiments; (2) low-level experiments revealing direct causality through understanding specific mechanisms. could focus on specific domains like sleep — understanding the range of how people sleep and what mechanisms cause blue light, eating, or exercise to impact it.
+
+this connects to the [[universal-data-capturer|universal data capturer]] for the data collection infrastructure. the [[always-on-ai-assistant|always-on AI assistant]] could passively gather the behavioral data needed for correlations.
+
+**spreadsheet evaluation:** originality 7/10, excitement 8/10, MVP 4-8 weeks. competitive landscape: QuantifyMe (MIT) and TummyTrials (UW) are research prototypes for n-of-1 self-experimentation, WHOOP/Oura show correlations but don't run controlled experiments, no consumer product provides unified causal testing across sleep/food/exercise. MVP could be a structured experiment planner + WHOOP data integration without making medical claims. tech depth 8/10, labeled [DO THIS] — stats + bio + data pipeline, controlled experiment design, causal inference, sensor integration, connects to sleep/biometric interests. blue ocean analysis: EMERGING market positioning with medium 2-month feasibility. the "N-of-1 Causal Inference Engine" variant in blue ocean analysis is the only idea rated true BLUE OCEAN with high 2-month feasibility.
+
+---
+
+## timeline
+
+- [2026-02-09] core concept — personalized impact testing framework
+- [2026-02-10] two approaches articulated — data collection vs mechanistic experiments
+- [2026-04-10] google sheets evaluation — 7/10 originality, 8/10 excitement, [DO THIS] tier, BLUE OCEAN for N-of-1 variant