Health Data & Intelligence. Fragmented data.Confident decisions.
We build the governed data foundation US health organizations need for AI, quality reporting, payer data exchange and revenue integrity, on AWS, Google Cloud or Azure, with privacy and security built in from day one. Then we put the answers in front of the people who decide.
A short call with a senior expert. No patient information needed.
Your sources
- EHR
- Claims
- Payer APIs (FHIR)
- Imaging & devices
- Registries
The governed foundation we build
- 1Ingest
- 2Standardize to FHIR and USCDI
- 3Match identity, de-identify
- 4Govern: access, lineage, audit
Decisions
- Dashboards
- Predictive models
- AI agents
- Quality & payer reporting
Built to the standards US health data runs on
- HL7 FHIR
- US Core
- USCDI
- SMART on FHIR
- TEFCA
- HL7 v2
- X12
- HIPAA Privacy & Security Rules
- De-identification (Safe Harbor, Expert Determination)
Watch 45 seconds
Five systems, no agreed numbers: how we build a governed data foundation
Read the film as text
- Five systems. No agreed numbers.Your data is everywhere, and nothing lines up.
- Why it matters$21B of administrative waste still to remove. Source: 2025 CAQH Index Electronic transactions already helped US healthcare avoid $258 billion in 2024.
- One governed foundation.Ingest. Standardize to FHIR and USCDI. Match identity, de-identify. Govern: access, lineage, audit. (built from the bottom up) On AWS, Google Cloud or Azure, with privacy built in from day one.
- Then the answers reach the decision.Dashboards. Predictive models. AI agents. Quality & payer reporting.
- During Health.Fragmented data. Confident decisions. Book a data readiness call, or Free readiness scorecard, at duringhealth.com.
What is changing around health data right now
Dated facts, each with its source. This part of the page is refreshed regularly; the rest is built to last.
70,000+
sites exchanging data on TEFCA
More than 474 million documents shared since go-live in December 2023. Payment and health care operations, including HEDIS reporting, are now authorized exchange purposes.
ASTP/ONC TEFCA update, Feb 19, 2026As of Oct 2026$21B
of administrative waste still to remove
Electronic transactions helped US healthcare avoid $258 billion in 2024; full automation of the rest could save about $21 billion more. Over half of health plans and a quarter of provider organizations already use AI in administrative work.
2025 CAQH Index, Feb 19, 2026As of Oct 2026Jan 1, 2027
payer FHIR APIs go live
CMS-0057-F requires Patient Access, Provider Access, Payer-to-Payer and Prior Authorization APIs: new, structured data streams that only organizations with a ready foundation can use.
Our CMS-0057-F analysisAs of Oct 20262030
accountable care for every Medicare beneficiary
CMS’s goal is that all Medicare beneficiaries with Parts A and B are in a care relationship with accountability for quality and total cost of care by 2030. That is a data-heavy way to be paid.
CMS Innovation Center strategyAs of Oct 2026
710
large health data breaches reported to HHS in 2025
About 36% happened at business associates, the vendors that hold data for providers and plans. Who builds and runs your data layer is a security decision, not just a technical one.
Every number links to its source.
What most teams get wrong about health data
Myth
“Our EHR already gives us the data we need.”
Fact
An EHR holds one slice. Claims, payer APIs, devices, imaging and registries live elsewhere, and quality and payment programs need them joined, matched to the right patient and defined the same way.
Myth
“AI is mostly a model problem.”
Fact
It is mostly a data problem. A model is only as complete, current and traceable as the data under it, and your board will ask where every answer came from.
Myth
“Interoperability is just a compliance deadline.”
Fact
It is a new supply of data. National exchange networks and payer APIs can feed care, quality and revenue work, if your foundation can take them in.
Four things you can hold us to.
Data engineering & platforms
Healthcare data lakes and warehouses, hybrid-cloud, built for scale and compliance.
What you get
- Target architecture on AWS, Google Cloud or Azure
- Pipelines from EHR, claims, FHIR and HL7 feeds
- PHI controls: role-based access, encryption, audit logging
- Automated data-quality checks and lineage
Predictive & decision intelligence
Models that forecast risk and surface the next best action.
What you get
- Models scoped to one decision, such as denial risk, no-shows or demand
- Validation on your own data before anyone relies on it
- Explanations and monitoring so reviewers can trust the output
- A human-review step for anything consequential
BI & visualization
Power BI and Tableau intelligence that leaders actually use.
What you get
- Executive and operational dashboards in Power BI or Tableau
- One written definition and one owner for each metric
- Role-based views for finance, HIM, quality and operations
- A data model your own team can extend
Multimodal data
Unify imaging, EHR, claims and biosignals into one analytical view.
What you get
- A unified patient and encounter model across EHR, claims, imaging and biosignals
- Identity matching and de-identification (Safe Harbor or Expert Determination)
- AI-ready datasets with provenance
- Mapping to USCDI and FHIR
Everything we do in health data and intelligence
Ten service areas, from the platform underneath to the dashboard on an executive’s screen. Most engagements start in one and grow into the others.
Data foundation & platforms
The governed base every other use case stands on.
- Cloud warehouse or lakehouse on AWS, Google Cloud or Azure
- Pipelines from HL7 v2, FHIR, X12 and flat files
- Patient identity matching and master data
- Data quality checks and lineage
Interoperability & data exchange
Getting data in and out of other organizations’ systems.
- FHIR APIs and bulk export
- HIE and TEFCA connectivity
- Payer data exchange and prior authorization APIs
- Mapping to USCDI and US Core
Clinical & quality analytics
Measuring care and closing the gaps.
- eCQM, digital quality measure and HEDIS pipelines
- Population health and risk stratification
- Care-gap and risk-adjustment analytics
- Safety, readmission and length-of-stay reporting
Revenue cycle & financial analytics
Protecting and speeding up the money.
- Denial analytics, root cause and prevention
- Charge-master and charge-capture analytics
- AR, collections and payer-mix realization
- Documentation and coding validation
Operational intelligence
Running the hospital or practice with fewer surprises.
- Capacity, throughput and scheduling analytics
- No-show and demand forecasting
- Staffing and labor-cost analytics
- Supply and utilization tracking
Value-based care & payer analytics
Data for contracts that pay for outcomes.
- Total cost of care and episode analytics
- ACO and shared-savings performance
- Utilization management and prior-authorization analytics
- Network and provider performance
Predictive & generative AI
Models and assistants that run on trustworthy data.
- Risk, demand and denial-prediction models
- Language models over notes and documents
- AI-ready datasets with provenance
- Model validation, monitoring and human review
Governance, privacy & security
The controls that make every other area safe to use.
- Role-based access, encryption and audit logging
- De-identification: Safe Harbor and Expert Determination
- Consent and sensitive-data handling
- Support for HIPAA security risk analysis
Research & real-world data
Datasets that stand up to scrutiny.
- Cohorts and registries from real-world data
- Imaging and multimodal datasets
- De-identified data for model training
- Reproducible, documented pipelines
BI & decision experience
Putting the answer in front of the person who decides.
- Power BI and Tableau dashboards
- A shared semantic layer, one definition per metric
- Executive scorecards and alerts
- Self-serve views for each role
One example: Healthcare Data Intelligence for the revenue cycle
A single application of Health Data & Intelligence + Intelligent Health Operations. The same rigor extends across every domain we work in — this is illustrative, not the limit.
Denial Analytics
- Cause categorization
- Root-cause & prevention
- Recovery optimization
CDM Analytics
- Structure validation
- Charge-to-code mapping
- Revenue-leakage detection
Clinical Doc & HIM
- Documentation completeness
- ICD / CPT / DRG validation
- Audit-risk indicators
Operational Workflow
- Access & registration accuracy
- Authorization performance
- Throughput & TAT
Financial Performance
- AR & collection efficiency
- Reimbursement variance
- Payer-mix realization
Predictive & Preventive
- Denial prediction
- Revenue-risk forecasting
- Preventable-error ID
How an engagement runs: start with one decision, not a platform
You see a working result on your own data before you commit to anything bigger.
Assess
Map your sources, data quality and the decisions that depend on them.
You get
Readiness report and a ranked list of use casesDesign
Agree the target architecture, the metric definitions and the governance.
You get
Architecture, metric dictionary and PHI control planProve
Build the foundation for one high-value use case and measure what it changes.
You get
A working pipeline, dashboard or model on your dataScale or transfer
Extend to the next use case, have us run it, or hand it over to your team.
You get
Runbooks, training and an optional build-operate-transfer path
Who we build this for
Health systems and hospitals
The problem
Reporting depends on spreadsheets and one-off extracts, and every department has a different number.
A good first project
Revenue-cycle and quality dashboards on one governed model.
Physician groups and MSOs
The problem
Several EHRs and no single view of performance, productivity or denials.
A good first project
Cross-practice denial and productivity analytics.
Payers and health plans
The problem
Clinical and claims data are hard to join for risk, quality and prior authorization.
A good first project
A quality-measure and payer-provider data exchange pipeline.
Health-tech companies
The problem
Their health-system customers want an auditable, AI-ready data layer before they will sign.
A good first project
FHIR-based ingestion and de-identified, traceable datasets.
Are you ready to use AI and payer data?
If you cannot tick most of these today, the foundation is the project, not the model.
- Every key metric has one written definition and one owner.
- You can trace any dashboard number back to its source system.
- Patients are matched across systems, with a measured error rate.
- Access to PHI is role-based, logged and reviewed.
- Claims and clinical data can be joined at patient and encounter level.
- You can receive FHIR data, such as payer APIs, without a custom project each time.
- Data quality is checked automatically, not discovered in a meeting.
- Anyone using AI on your data is working under an approved policy.
Three questions to ask your team on Monday
- 01
Which three numbers do we report to the board, and can we show where each one comes from?
- 02
If a payer’s API sent us structured clinical data tomorrow, where would it land and who would use it?
- 03
How many places hold a copy of PHI that we are not actively governing?
If the answers are slow or different in every room, that is where we start.
Check our thinking before you call us
We publish how we read the rules and give away the tools. Judge the depth for yourself.
- Article
CMS-0057-F and prior authorization: what providers must be ready for
The four FHIR APIs, the dates and a 90-day checklist.
- Article
Medicare EDI and HIPAA transactions explained
Every transaction in the life of a claim, and the latest changes.
- Tool
HIM technology readiness scorecard
Score your readiness in 15 minutes. Nothing is stored.
- Tool
NCCI edit helper
A plain-English answer to a Medicare coding denial.
Which cloud do you build on?
AWS, Google Cloud or Azure. We choose with you, based on where your EHR, security team and contracts already sit, so you are not locked into one vendor’s product.
How do you handle our PHI?
Only under a written agreement, including a business associate agreement where HIPAA applies, signed before any PHI is shared. We design for the minimum necessary: role-based access, encryption and audit logging, and we work on de-identified data wherever the use case allows. Read our trust center for how we work.
We already have a data team and an EHR vendor. Why add you?
We work alongside them. We take on the cross-source foundation, the metric definitions and the first use case, so your team keeps its time for the work only it can do.
How do you de-identify data?
HIPAA recognizes two methods. Safe Harbor removes 18 types of identifiers. Expert Determination has a qualified expert document that the risk of identifying anyone is very small. We help you choose the one that fits the use case and keep the evidence.
How long until we see value?
We start with a single, defined use case so you can judge a real result before committing further. Timing depends on data access and scope, and we set the first milestone in the assessment.
Is this about AI or about data?
Data first. AI projects stall on incomplete, inconsistent or untraceable data. Once the foundation is governed, AI, analytics and quality reporting all get cheaper and safer to run.
What does an engagement cost?
It depends on scope. We begin with a defined first use case and a clear deliverable, so you can see what you are buying before you commit to a larger program.
Start with one decision. Build the foundation behind it.
Tell us the decision you wish you could make with confidence. We will tell you what data it needs, what is in the way and how we would start.
Please do not send patient information. Describe the situation without identifying anyone.