Shifting digestion and recurring mild nausea or fatigue wear people down and push them to quit early.
Medication gets people started. Staying with it is a daily, behavioral problem. Here is the design thesis behind Nishi, in plain language, and the working paper it comes from.
GLP-1 medications regulate appetite and blood sugar well. But the paper's starting point is a sobering one: a large share of people, by some estimates up to half, stop within the first year. Rarely because the drug failed. Usually because daily life got in the way.
Shifting digestion and recurring mild nausea or fatigue wear people down and push them to quit early.
Too little protein during rapid weight loss costs lean muscle and slows metabolism, the opposite of the goal.
Logging every meal and symptom into a blank prompt is friction. People abandon the tracking long before the plan.
The idea at the center of the paper: flip the app from reactive to proactive. Instead of waiting for you to type, the system watches context in the background and surfaces one low-friction, useful step at a time.
Reads overnight recovery and where you are in your dose, then suggests a gentle, protein-forward breakfast that is easy on nausea.
Tracks protein and energy as the day builds and nudges a small adjustment before the afternoon dip sets in.
Checks what is in your kitchen against your targets, queues gentle recipes, and fills in what is missing for your grocery cart.
The global rollout of GLP-1 receptor agonists is a major advance in chronic disease management, but it shares one operational failure point across patients and health systems: high discontinuation. The paper sets out an active biometric observation architecture designed to run alongside treatment, replacing passive, high-friction manual logging with a proactive observation loop of background telemetry, real-time symptom management, and automated dietary execution.
This is a working paper that lays out a design thesis. It argues a model for supporting adherence and makes an economic case for it. It is not a clinical trial, and it does not claim proven medical outcomes. Nishi is a wellness companion, not a medical service.
Start with your targets. A 60-second quiz, no signup needed.
Find your targets