The Challenge

SUBMITTING DATA MEANT SUBMITTING IT INTO A BLACK BOX

Public health depends on getting critical data from the ground to the people making decisions with it, fast. But health authorities across the country, state, local, tribal, and territorial, each had their own disconnected system for submitting that data to a federal agency, with no single, reliable channel to rely on.

Once submitted, there was no way to know if the data had arrived, or if it was even correct. Partners often waited days or weeks just to find out, and that delay slowed down analysts and decision-makers exactly when speed mattered most.

What We Built

PARTNERS KNOW THE MOMENT DATA ARRIVES

Health authorities submitting data now get real-time validation of both metadata and content, for critical formats like HL7, CSV, and FHIR, the moment they submit. Issues surface as data enters the system instead of downstream, so partners can correct and resubmit right away instead of waiting days or weeks to find out something went wrong.

A DROPPED CONNECTION DOESN'T MEAN STARTING OVER

A health authority submitting a 20-gigabyte dataset over an unstable connection doesn't have to start over if the connection drops, the upload resumes automatically. That same reliability held up under real-world conditions, including sustained ingestion at pandemic-level surge capacity.

BUILT TO WORK ANYWHERE

Partners never had to worry about the platform being tied to one infrastructure provider, the exchange runs cloud-agnostic, and proved it with a successful migration from Azure to AWS without losing functionality or disrupting anyone using it. That flexibility also cut operating costs by more than 80 percent.

THE SYSTEM NEVER GOES OFFLINE

Partners submitting time-sensitive public health data never lose access when an update goes out. New versions come online before old ones are switched off, so nothing is ever offline and nothing is ever interrupted.

Outcomes

1 gig pr minute
data processed from state, local, tribal, and territorial authorities
83%
cost reduction, lowering monthly expenses from $12,000 to $2,000
140 million
objects processed in 19 hours
16
Federal program partners integrated into the platform
<1 hour
processing time for each dataset