Coach in chat
Ask what to do today, log a set, send feedback, or talk through a meal. The assistant uses your stored context instead of starting from a blank prompt.
neurologyYasker / Personal health intelligence
AI health coach with memory
Yasker learns your body over time by connecting wearable signals, workouts, meals, labs, and daily feedback into one adaptive coaching loop for performance, body composition, and long-term health.
Product
Less dashboard. More decision.
Most health apps ask you to interpret charts. Yasker turns the data into the next safe action: what to train, how hard to push, what to eat, what to track, and when to recover.
Wearables
Sleep, HRV, resting heart rate, readiness, activity, and training load.
Training
Set logs, RPE, soreness, pain signals, performance trends, and form notes.
Nutrition
Meals, protein targets, food photos, adherence, and daily context.
Labs
Blood markers, prescriptions, diagnoses, dates, provenance, and guardrails.
Memory with structure
Important facts become durable records: workouts, metrics, labs, meals, insights, and coaching decisions can be inspected and improved.
Readable audit trail
Every coach request, AI response, tool call, and error can be reviewed so the product gets better instead of becoming an opaque black box.
System
The loop is the product.
Yasker is designed around small daily decisions that compound. It captures what happened, interprets what matters, and adjusts the next recommendation.
Capture the day
Wearable data, workouts, meals, labs, and subjective feedback enter one personal data layer.
Separate signal from noise
The coach reads recent context, looks for patterns, and stays honest when evidence is weak.
Adapt the next action
Tomorrow's plan changes based on sleep, soreness, performance, nutrition, stress, and life.
Yasker Today: recovery is low but strength trend is stable. Recommendation: • keep the session, reduce top sets by 5-8% • cap intensity at RPE 7 • add 20 minutes zone 2 only if soreness stays mild Logged context: sleep 6h 10m · HRV down · squat +2.5kg last week · shoulder pain none
Principles
Built for serious personal data.
Healthspan coaching only works when the product is careful with uncertainty, provenance, privacy, and user trust.
Doctor-in-the-loop
Yasker supports decisions and education. It does not diagnose disease or replace clinical care.
Data before opinion
Recommendations should cite the user's actual records whenever possible: workouts, recovery, labs, and feedback.
Inspectable memory
The coach should be auditable. Stored facts, tool calls, and responses must be readable by humans.
Neutral interface
The system should feel calm, precise, and durable, with color used for action and meaning, not decoration.
MVP