Strengthen Health Cost Prediction
at the Patient & Population Level

PRECISION
More accurate health spend forecasts for payers and risk-bearing care providers. Lumiata generates high-performance machine learning cost models, based on thousands of data points - clinical, care provider, utilization, and system engagement.

PERSONALIZATION
Spend forecasts tailored to patients and how they receive care at the local level. Lumiata differentiates its patient-level predictions based on diverse criteria such as clinical history, regional characteristics, and market factors.

FOCUS
Risk and cost predictions based on financial and disease models. Lumiata directly models cost predictions using patient data. Data spans the entirety of historical information which is multi-fold, compared to legacy approaches.

Health Analytics for Managing Risk and Cost​

HEALTH PLANS

​Improve alignment between risk and sales by pinpointing predicted costs ​for underwriting

RISK-BEARING CARE PROVIDERS
​Better anticipate utilization at the individual and population level to support contracting

EMPLOYERS
​
​Predict trend across employee health spend; identify changes in risk & carrier engagement

We transform data into insights.
Slide the < > below to see how we do more with data.
​

Stage 1
Medical Data
to Patient Records

Patient events (claims, labs, drugs and procedures) are structured into FHIR, a standardized format for healthcare data.

Data Engineering

Stage 2
Patient Records
​to Patient Journeys

Transformations of multiple events form features that define patient journeys, mapping historic patient data to present health states.

Medical Data Enrichment

​Stage 3
Patient Journeys
​to Lumiata Model

Through machine learning, Artificial Intelligence (AI) informs model insights.

Pattern Discovery

​Stage 4
Lumiata Model
​to Lumiata Insights

With Lumiata Insights, at both the patient and population level, cost predictions for health spend forecasts can be accurately made in a timely fashion.

Prediction

We transform data into insights.

Slide the < > below
​to see how we do more with data.​​

Stage 1
Medical Data
​to Patient Records

Patient events (claims, labs, drugs and procedures) are structured into FHIR, a standardized format for healthcare data.

Stage 2
Patient Records
​to Patient Journeys

Transformations of multiple events form features that define patient journeys, mapping historic patient data to present health states.

​Stage 3
Patient Journeys
​to Lumiata Model

Through machine learning, Artificial Intelligence (AI) informs model insights.

​Stage 4
Lumiata Model
​to Lumiata Insights

With Lumiata Insights, at both the patient and population level, cost predictions for health spend forecasts can be accurately made in a timely fashion.

​We draw from a great depth of medical knowledge.

MILLION

PATIENT RECORDS

MILLION

CONNECTIONS BETWEEN
​MEDICAL CONCEPTS

THOUSAND

PHYSICIAN CURATION HOURS

TERABYTES

UNSTRUCTURED DATA

MILLION

ARTICLES MINED FROM PUBMED

MILLION

MEDICAL GRAPH EDGES