Health data, at a sguardo.

Sguardo aggregates, normalizes, and structures health records from any source — so providers can measure what matters and individuals can finally see the full picture.

Built on OMOP CDM and FHIR. On-prem or cloud. HIPAA-compliant. No proprietary lock-in.

Limited Private Beta

Sguardo is currently in limited private beta. Request an invite and we'll get you set up personally.

For Providers & ACOs For Individuals

Why health data doesn't work yet

Fragmented

A typical Medicare beneficiary sees seven different providers a year. Each one records into a different system. The cardiologist can't see what the endocrinologist ordered last week.

Inconsistent

Your pharmacy calls it "metformin HCl 500mg." Your hospital records "Glucophage 500 mg oral." Your insurer says "metformin 500mg tablet." Same drug — but no system treats them that way.

Locked Away

The 21st Century Cures Act gave patients the legal right to their data. In practice, most people get a PDF they can't read. Providers get marginally better access, but cross-system queries remain nearly impossible.

Privacy & Safety First

AI that runs on your infrastructure, not someone else's cloud.

While every other health tech company races to send your patient data to OpenAI or Anthropic, Sguardo keeps AI local. We use small, purpose-built models for language understanding — powered by Predicato — that run entirely within your infrastructure. No third-party AI services receive your PHI. You control the de-identification level and can inspect exactly what goes over the network.

🚫

No third-party AI touches your data

Sguardo uses small, purpose-built models (Qwen3-0.6B for embeddings, SmolLM2 for text, GLiNER for NER) that run locally in the application via Predicato. Nothing is sent to OpenAI, Google, Anthropic, or any third-party AI provider.

🏠

Runs on your infrastructure

On-prem or your cloud. AI models run locally via Predicato. Or use our hosted service with full control over de-identification and full transparency into what data is sent. Vocabulary lookups, reference ranges, and scoring are all embedded locally.

👁

Open and auditable

The ETL pipeline is inspectable. The care gap logic shows its numerator and denominator. The HCC scoring algorithm is documented math. You don't have to trust a vendor — you can read the code.

🔒

We don't collect your data

Sound Prediction's business model does not depend on harvesting health information. We have no incentive to collect, aggregate, or monetize your patient records — because we don't.

What Sguardo does about it

A single platform that handles the entire pipeline from raw data to clinical insight. Deterministic algorithms, so every transformation is reproducible and auditable.

Aggregate

SMART on FHIR for EHRs, CMS DPC and Blue Button 2.0 for Medicare data, Apple Health and Google Fit for patient-generated data, and standard connectors for legacy databases.

Discover

Sguardo scans source databases using a methodology inspired by OHDSI's WhiteRabbit, profiles value distributions, and measures completeness — giving you a structural map without exposing PHI.

Normalize

Map source data to the OMOP Common Data Model — the same standard used by the OHDSI network at 400+ institutions. Deterministic concept matching, auditable output.

Analyze

HEDIS quality measures, HCC risk coding, care gap identification, lab reference ranges, population health dashboards, and financial decision support.

Two products. One platform.

Sguardo for Providers

For FFS and VBC practices alike. Sguardo connects your systems, normalizes everything to OMOP CDM, and delivers quality measures, HCC coding, care gap identification, population health dashboards, and financial tools for rate optimization and contract negotiation.

Learn More

Sguardo for You

Connect MyChart, Medicare Blue Button, Apple Health, and Google Fit in minutes. Get a unified health dashboard with interpreted lab results, medication tracking with GoodRx pricing, conditions in plain English, and true data ownership. Try demo mode instantly.