Skip to content

Who we serve › Public health agencies

You are accountable for outcomes you do not control.

Gaine works across the systems your network already runs. No organization in it is asked to move, adopt, or conform. Gaine resolves the people, the organizations serving them and the services delivered, wherever those records sit, and hands back the crosswalk, the lineage and the submission in formats you keep.

Request a demoHow the record reaches your systems

Where this stops

Some contributors are on paper. Here is what happens then.

An organization with a case management system exports on a schedule it already keeps. One with a spreadsheet sends the spreadsheet. One working on paper sends what it has, and somebody keys it — and that somebody should be named in the contract before anyone signs it, because unpaid data entry is how these programs quietly fail.

What we will not do is make participation conditional on a partner changing systems. Not because it would be technically hard, but because you cannot compel it and every proposal that assumes you can is a proposal you cannot execute.

The honest constraint: the fewer structured feeds a network has, the more of the work is manual at the edges. Gaine reduces what has to be reconciled centrally. It does not make a paper record arrive by itself.

What you end up holding

The crosswalk is yours, and it keeps working without us.

Resolution produces an artifact, not a subscription: a crosswalk mapping every contributing system’s identifier to the resolved one, exported in a standard format on a schedule you control. Flat files, a replicated store, something a reporting tool can sit on without a proprietary layer in between.

If your agreement with Gaine ends, the crosswalk you already hold does not stop being true. Historic submissions stay reconstructable and the identifiers you reported against stay valid. What you lose is future resolution, not the record of what was resolved.

This is stated plainly because this sector buys with grant money and has been left stranded before. Dependency is a bigger objection here than cost.

The calendar, not the database

The problem is a submission date, not data quality.

Recurring submissions to the state arrive on weekly, monthly and quarterly cycles, drawn from a care coordination platform, a CRM, an accounting system, an eligibility file, a health information exchange and several spreadsheets. In most networks one person assembles them by hand.

Gaine produces the reconciled figures in the format each submission requires, with the reconciliation attached — so the work before a deadline is review, not assembly. Where a figure has changed since the last cycle, the change carries the reason it changed.

Worked example

One person, three identifiers, three times counted.

The same person is screened at a food pantry, referred by a clinic, and appears on an eligibility file from the state. Three systems, three identifiers, no shared key. In the quarterly figures they are three people served.

That is not a rounding error. It inflates the denominator on every rate you report, it makes the closed-loop percentage unprovable, and when someone eventually reconciles it, the correction lands as a restatement on a figure a funder has already acted on.

Resolving the three into one person, with the evidence for why they were joined and the ability to separate them again if the match was wrong, is the precondition for every number downstream of it.

Defensible months later

Every figure traces back to the records that produced it.

Field-level lineage on each reported number: which contributing systems supplied it, which records resolved together, which rule decided a conflict, and what the figure was on the date it was submitted rather than what it would be if recalculated today.

A funder query arriving eight months after a submission is answered from the record rather than from memory. That is the difference between a question and an audit finding.

Not just people

The hard matches are organizations and services.

Every vendor has a good number for matching people. The two that break a reporting pipeline are quieter.

  • The same organization under four names. A community organization appears with its legal name in a contract, a trading name in a referral platform, an abbreviation in a spreadsheet and a site name in an eligibility file. Gaine resolves them and keeps the relationships between parent, program and site.
  • Service categories no two funders define alike. One taxonomy for the waiver, another for a grant, a third for the state. Gaine holds them side by side and maps between them, so a service delivered once is reported correctly in each.
  • Multiple referral platforms in one community. Partners adopt different platforms and none resolves identity across the others. Gaine sits beneath them all.

Who runs it

Built to be operated by one person, part-time.

Most networks have one technical person, or a consultant retained a few days a month. Anything that needs a specialist to keep running is a liability here rather than a feature, and anything that needs a vendor engineer for a routine change is a queue you do not control.

Match rules, contributor onboarding and submission formats are configured and reviewed without writing code, and the review queue is sized so a part-time steward can clear it. Where a decision genuinely needs judgment, it surfaces as a small number of exceptions rather than a backlog.

Start with one submission.

Pick the report that costs you the most time and the most argument. We will show you what resolving it looks like before you commit to anything wider.

Request a demoTalk to an architect