Skip to content
HALEMedical and Data Sciences

Software and data for the nation’s health.

Hale Medical and Data Sciences brings test-driven software, medical data sciences, predictive analytics and multimodal AI to the government agencies that safeguard American health — engineered to federal standards, with clinical expertise in-house.

Explore capabilitiesSchedule a briefing

Frequent mental distress among adults, by county

Age-adjusted prevalence, 2022.

Loading county estimates…

Each bar is a group of counties. Hover a bar to find them on the map; click a county to zoom to its state.

Source: CDC PLACES county estimates, published 2024, from the 2022 Behavioral Risk Factor Surveillance System. Veterans: VA’s VetPop2023 county estimates, 30 September 2023. Connecticut reports by planning region and is not drawn; territories are not shown.

Software & Data EngineeringNAICS 541511 (primary), 541512, 518210

Connecting systems that were never designed to talk.

Custom applications and health data pipelines, interoperability across FHIR, HL7 and X12, and secure hosted environments for sensitive data — built to federal security and accessibility standards.

A day of health-data traffic through an interface engine

Messages a day by route. Hover a system, or tap it, to trace everything that flows through it.

  • Message routes
  • Held for review
EHR admissionsHL7v2 ADT, 184k/dayClinical recordsFHIR R4, 96k/dayClaimsX12 837, 41k/dayLab resultsHL7v2 ORU, 68k/dayImmunizationsHL7v2 VXU, 12k/dayInterface engine401k a dayTerminology services302k a dayHeld for review13k a dayClinical decision supportAlerts and order sets, 152k/dayQuality reportingeCQM measures, 54k/dayPublic-health registriesLabs and immunizations, 38k/dayResearch enclaveOMOP CDM, 144k/day
Routes as a table
Messages a day by route
FromToA day
EHR admissionsInterface engine184k
Clinical recordsInterface engine96k
ClaimsInterface engine41k
Lab resultsInterface engine68k
ImmunizationsInterface engine12k
Interface engineTerminology services302k
Interface engineResearch enclave86k
Interface engineHeld for review13k
Terminology servicesClinical decision support152k
Terminology servicesQuality reporting54k
Terminology servicesPublic-health registries38k
Terminology servicesResearch enclave58k

The flow diagram and the patient cohort on this page are illustrative product concepts, not screenshots of delivered systems. Their volumes and patients are invented.

Software and data engineering capabilities

Health Research & AnalyticsNAICS 541715, 541720Clinical & Advisory ServicesNAICS 621112, 541690

Who a risk model flags, and whether the program helped.

Predictive models and analysis on data agencies already hold, with physician-led clinical judgment on what an outcome measure actually captures.

700 hospital discharges, placed by predicted 30-day readmission risk

Each dot is one patient. Drag the cutoff to choose who gets transition care, then split the cohort by program.

  • Readmitted within 30 days
  • Not readmitted
  • Faded: below the cutoff
0%10%20%30%40%50%60%70%Predicted 30-day readmission riskCutoff 20%
Patients flagged
207
patients flagged, 30% of all discharges
Readmissions caught
54%
of readmissions caught (44 of 82)
Flagged patients readmitted
21%
of flagged patients were readmitted
Research and analytics capabilitiesClinical and advisory capabilities

Bring us a hard problem.

A system that has to talk to others it was never designed for, data that has to answer a question, or a program that needs a physician’s judgment — we start with what would count as done.

Start the conversation