Your patients’ data is scattered. Glance brings it together.

Every provider sees a fragment. We build the infrastructure that assembles the full picture — so clinicians can act on complete information, ACOs can close care gaps before they cost money, and patients finally understand their own health.

Sound Prediction -- Health data infrastructure for sound decisions

Recognized by

CMS AI Health Innovations Challenge Semifinalists
CMS Data at the Point of Care Partner Vendor
Published in ICLR, AISTATS, MLHC & PLOS
NIH-affiliated research team
Health Data Platform

Glance

Your EHR only knows what happens inside your walls. Claims arrive weeks late. External records come by fax. Glance fixes the underlying problem: it pulls records from every source, normalizes everything to OMOP CDM, and surfaces the analytics you actually need — with purpose-built local AI, no third-party data sharing, and no PHI leaving your infrastructure.

For providers and ACOs: HCC risk coding, care gap identification, quality measure computation, and financial analytics — built on CMS DPC and SMART on FHIR. For individuals: a unified health record with lab interpretation, medication tracking, cost transparency, and local-first privacy.

Privacy & safety first. Purpose-built local models — no third-party AI services touch your data. Every algorithm is auditable and runs entirely on your servers.

glance.local
Data Sources
8
Patients
45.2K
ETL & Data Integration
Schema discovery · OMOP mapping · SQL generation
CMS Data Services
DPC · Blue Button 2.0 · Quality measures
Open Source

Predicato

Predicato is our open-source temporal knowledge graph framework. Extract entities and relationships from documents using local ML models, build knowledge graphs that evolve over time, and query across your entire knowledge base — all without external API calls.

Go library with Python client. Embedded databases and local ML via Rust FFI. Bi-temporal modeling, hybrid search (semantic + keyword + graph traversal), entity resolution, and community detection. No API keys, no vendor lock-in, no recurring costs.

// Extract knowledge from a payer contract
episode := client.AddEpisode(ctx, contract,
  WithExtraction(true),
  WithExtendedTriples(true))
// → 47 entities, 83 relationships, 12 rules
// Search with hybrid retrieval
results := client.Search(ctx, "prior auth MRI")
// → IF CPT 70553 AND outpatient,
//   THEN prior auth required
//   Source: contract.pdf, page 14

From our team

Perspectives on health data, clinical technology, and why we build the way we build.

November 05, 2024

Paper forms don't have to be dumb

There’s a recurring debate in nursing that tends to generate more heat than light: paper versus digital documentation...

Read more →
July 22, 2024

The hidden cost of missed Annual Wellness Visits

The Annual Wellness Visit (AWV) is one of the most consistently under-utilized preventive care benefits in Medicare. ...

Read more →
March 15, 2024

Why we built Glance on OMOP CDM

When we started designing Glance, we had a choice: build a proprietary data model optimized for our specific use case...

Read more →
January 19, 2023

Why we don't use LLMs for clinical data transformations

When we started building Glance, one of the earliest decisions we made — and one we’ve held to ever since — was to ke...

Read more →

Key partners


Mederrata Research Inc. logo
Mederrata Research is a non-profit working on reducing medical errors.