07 · AI

AI, Analytics & MLOps

Dataset management, model registry, prediction & personalization.

Models in registry
38
Training jobs (24h)
142
Datasets
1.6M
samples
Predictions / sec
3,240

Training run · glucose-est-v4.2

epoch 38/40
Train loss
0.082
Val loss
0.094
RMSE
14.8
MAE
11.4

Holistic health score

Wellness
84
Cardiac88
Metabolic72
Respiratory94
Immune81

Model registry

ModelTaskPerformanceTrainedStatus
glucose-est-v4.2Glucose regressionMARD 8.2%Jun 12Production
af-detector-v2.0AF classificationF1 0.94May 22Production
stress-index-v3.0Stress regressionRMSE 0.12Jun 26Beta
sleep-stage-v1.4Sleep stagingAcc 0.86Apr 02Production
hba1c-pred-v1.1HbA1c forecastMAE 0.3%Jun 28Shadow

Dataset · training

1284.0k
samples
Stratified by age, gender, ethnicity. Drift score 0.04 (stable).

Dataset · validation

160.5k
samples
Stratified by age, gender, ethnicity. Drift score 0.04 (stable).

Dataset · test

160.5k
samples
Stratified by age, gender, ethnicity. Drift score 0.04 (stable).

Prediction engines

AF risk · 24h
3.2%
Glucose · next 2h
142 mg/dL
Stress spike · today
likely 17:00
Sleep onset latency
14 min