Fractional Healthcare Data Engineering

Build The Healthcare Data Platform You Actually Own

We embed senior healthcare data engineers with your team to build your platform in-house on the open-source Tuva health data model instead of renting one from a vendor. When the engagement ends, the platform, the code, and the know-how stay with you.

  • You own everything we build

  • Built on the open-source Tuva health data model

  • PHI never leaves your cloud

Tuva Project Maintainers

We maintain part of the open-source Tuva Project itself: the service category and encounter groupers that thousands of downstream measures depend on.

How We Work

How We Make Healthcare Data Trustworthy

We work inside your environment, define the universe early, reconcile the measures that matter, and look at outputs with your team before they reach decision-makers.

Universe, Reconcile, Look keeps the work simple to explain and hard to misunderstand.

  1. U

    Universe of Data

    Every insight depends on the data you are actually looking at. We define scope up front: what sources are included, what is not, and what questions the analysis can and cannot answer.

    Claims data is great for cost and utilization, but it will not answer questions about labs unless those clinical feeds are intentionally included.

  2. R

    Reconcile Your Data

    After ETL and transformations, we do not just hope the numbers are right. We prove it by reconciling key measures back to legacy systems and source-of-truth reports.

    We compare counts, dollars, and other critical metrics so teams can see that before and after totals match expectations.

  3. L

    Look at the Data

    Analyst review is where results get validated. We sanity check outputs, investigate outliers, and pressure-test measures against what is typical for similar populations.

    Automated checks catch structure issues, but human eyes find what rules miss so stakeholders trust what the data is saying.

Ways We Help

How We Partner With Your Team

We're the team you call when you want to own your analytics stack instead of renting it, whether that's building the platform, hardening the metrics, or adding senior hands to your own team.

Build The Platform You Keep

We design and build a modern healthcare data platform in your cloud, on the open-source Tuva health data model, unifying claims, clinical, and operational data into a reconciled foundation your team runs long after we're gone.

Make Metrics Decision-Ready

We define the measures that matter, document lineage, and reconcile outputs back to trusted sources so stakeholders can act on the numbers.

Fractional Senior Engineering

Add senior healthcare data engineers to your team for a fraction of a full hire. We join your planning, ship in your repos, and document as we go. No junior bench, no lock-in, no black boxes.

Products

Tools We've Built Along The Way

Working software that grew out of real delivery work. These tools show how we build: transparent logic, warehouse-native execution, and no PHI leaving your environment.

EMPI

EMPI Engine + Workbench

Transparent patient identity resolution with warehouse-native matching, analyst review, and deterministic golden records.

  • Warehouse-native probabilistic linkage
  • Manual review workbench and visual config editor
  • No external PHI movement or SaaS runtime

Predictive Models

Predictive Models

dbt-native predictive modeling that trains and scores directly in your warehouse using your own claims history and target policy.

  • Train on your own data, not benchmark cohorts
  • Registry-aware model reuse and reproducible outputs
  • Standard warehouse tables for predictions and evaluation metrics

Spotlight DQ

Spotlight DQ

Healthcare data quality monitoring for reconciliation, completeness, and terminology validity with PHI-safe aggregate outputs.

  • Claims and enrollment reconciliation by month and plan
  • Field-level completeness with applicability rules
  • Terminology validation with trendable official-run history

The Team

Who You'll Work With

Illuminate Health is a boutique healthcare data engineering partner for organizations working in value-based care and population health. We're not a big firm with a rotating bench; you work directly with the people on this page.

We've spent our careers inside healthcare data, across consulting firms, health tech vendors, and provider organizations, and we still genuinely like the work. We contribute back to the open-source Tuva Project as a certified partner and as maintainers of its service category and encounter groupers.

We work like embedded engineers: we join your planning, ship in small increments, demo often, and write things down so your team can run what we build without us.

Our default delivery model is URL: Universe, Reconcile, Look. Define scope, prove the numbers, then validate results with human judgment.

Headshot of Tom Sherwood

Tom Sherwood

Partner & Co-Founder

Tom leads analytics engineering and product delivery in value-based care: the measure logic, the reconciliation, and the demos that make the numbers make sense to the people who use them.

Headshot of Brad Montierth

Brad Montierth

Partner & Co-Founder

Brad leads data architecture and engineering: the cloud platform, the pipelines, and the Tuva-based data model everything else stands on.

Headshot of Lincoln Haycock

Lincoln Haycock

Executive Advisor

Lincoln was Chief Analytics Officer at Castell, Intermountain Health's population health company, where he built the enterprise analytics function and directed the platform strategy behind one of the top-performing value-based care organizations in the country.

Contact

Start With A Practical First Conversation

We'll start with a 30-minute review of your platform, metrics, and delivery constraints, then map a practical first milestone.