Science & Research

Built to turn everyday health data into questions worth studying.

MATILDA brings together longitudinal signals and personal context to explore how daily routines relate to health over time. Its research framework centers transparency, patient relevance and evidence that can be examined.

Longitudinal signalsPersonal contextChanging patternsQuestions for review

A whole-person research lens

The signal matters. The circumstances around it matter too.

A glucose reading, meal, night of sleep or period of activity can be informative on its own. Studying how those events relate across time can provide a more useful view of everyday health.

Longitudinal

Look beyond one moment.

Patterns become easier to examine when readings are viewed across days, weeks and changing routines.

Contextual

Keep daily life in the record.

Meals, activity, sleep, medication, schedules and observations help frame what the numbers may mean.

Personal

Study change against the individual.

Patient-specific baselines provide a more relevant reference than treating every person as an average.

Research framework

A disciplined path from observation to evidence.

MATILDA’s research approach separates what the platform observes, what its model suggests and what can be supported through formal evaluation. That distinction helps keep product language aligned with the strength of the available evidence.

The website does not claim diagnosis, treatment efficacy or guaranteed health outcomes. Future claims should be tied to the appropriate evidence and review.
01

Define the questionIdentify the behavior, relationship or user need being examined.
02

Specify the signalsDocument the data sources, context and time window relevant to the question.
03

Evaluate the patternAssess repeatability, usefulness and potential sources of bias or error.
04

Validate the interpretationCompare model output with appropriate expert review and real-world evidence.
05

Communicate the limitsState what the result supports, what remains uncertain and what requires further study.
Questions guiding the work

Research begins with the problems people actually face.

MATILDA’s connected model creates opportunities to study how context can improve understanding, engagement and the conversations surrounding care.

01Can personal baselines clarify change?Explore whether comparing a user with their own history makes patterns easier to interpret.
02Which kinds of context are most useful?Examine how meals, activity, sleep, medication and personal observations contribute to understanding.
03Can explanations support engagement?Study whether clear, contextual explanations help users recognize practical next steps.
04Can connected histories improve conversations?Assess how longitudinal records support preparation and communication between patients and care teams.

Evidence
review

AccuracyRelevanceConsistencyFairnessClarity

Model evaluation

Useful intelligence must be tested from more than one angle.

MATILDA’s in-house machine-learning model was designed to learn from each user’s authorized data and adapt its internal understanding over time. Evaluating that system requires more than measuring technical performance alone.

Research must also examine whether outputs are understandable, relevant to the individual, consistent over time and appropriate for the decisions they are meant to support.

Responsible collaboration

Built for work that crosses disciplines.

Meaningful evaluation can involve patients, clinicians, behavioral scientists, data scientists, health-system partners and technology collaborators. Each perspective helps test a different part of the experience.

MATILDA welcomes conversations with organizations interested in responsible digital-health research, pilot design and real-world evaluation.

Patients and caregiversUsability, relevance, burden and lived experience.
Clinicians and care teamsInterpretability, workflow fit and clinical context.
ResearchersStudy design, measurement, analysis and transparent reporting.
Technology partnersInfrastructure, interoperability and responsible model operations.
What MATILDA supports

Evidence should strengthen understanding without overstating certainty.

Supports

Pattern exploration

Organizes connected information so relationships and changes can be reviewed more clearly.

Supports

Informed discussion

Helps people and authorized care teams prepare questions using a more complete history.

Does not replace

Medical judgment

MATILDA does not diagnose, prescribe treatment, provide emergency monitoring or replace qualified care.

Advance the evidence

Help study what becomes possible when health data keeps its context.

Connect with MATILDA about research, clinical collaboration, pilot opportunities and responsible technology development.

Scroll to Top