Prove delivery: every file lands in an inbox the moment it is seen, with duplicate detection, manual upload, and batch history.
Did the file arrive, was it readable, is it correct.
Every other number in Alara Population Health depends on data getting here and being read correctly. This module answers three separate questions, and shows each one separately, because they fail independently.
Data Management
Did the file arrive, was it readable, is it correct. Schema alerts, column mapping, pipeline scheduling, and an A–F data quality grade.
What it is for
What Data Management does
Read it safely: when a payer adds, renames, or drops a column, the pipeline stops and asks rather than guessing; file types are versioned and delimiters detected automatically.
Judge correctness: a continuous check engine rolls into an A to F grade and a 0 to 100 health score, findings name a steward and link to the source file, and datasets that fail badly enough are withheld rather than served.
Hold sources to their promises: a delivery calendar of what each source owes and whether it landed, and the payer contract terms each feed is scored against.
The screens
The screens, in the order you would use them
- 1
Inbox
What has been delivered: files in storage, manual upload, and batch history, with tiles for total, processed, awaiting schema approval, and unmapped.
- 2
Schema
What the product expects a file to look like: schema alerts, versioned file types, and the approval step when a layout changes.
- 3
Pipeline
What happened when a file was read: a two-panel column-mapping editor with sample values, the processing log, batch errors by file and type, and a scheduler with cadence presets.
- 4
Quality
Whether the contents hold up: the check engine, the report card, seasonality-aware anomaly detection, freshness service levels per file type, and identity resolution with ambiguous pairs queued for a human decision.
- 5
Sources
What was promised: the delivery calendar, and payer contracts with allocation rate, loss-ratio target, profit share, and the HEDIS measure set the contract is scored against.
Honest numbers
What the figures count
Every figure on every screen has a definition behind it. These are the ones this module runs on.
- Total Files and recognized files
- A file appears the moment it is seen, before anything has been read out of it. The recognized count beneath the tile is the one that matters.
- Total Rows and Complete
- Complete means the row was usable. Whether its contents are correct is a different question and a different screen.
- Data quality grade
- An A to F letter and a 0 to 100 score from automated checks. When a check fails badly enough, the affected numbers elsewhere in the product are withheld rather than shown wrong.
What can go wrong
What this module will not let you misread
Reading detected as usable
A file in the inbox has been seen, not read. Pipeline tells you what was actually read; Quality tells you whether it holds up.
Reading a stuck completion share as slow rather than stopped
Early in a run a low share is just work outstanding. A share that stops moving means the remainder is failing, and Batch Errors on the same screen shows what.
Assuming a complete row is a correct row
Completion means the row was usable. Quality is a different question.
Where the data comes from
Payer enrollment rosters and eligibility, payer-identified HCC and quality gap files and the year-end risk report, CMS encounter, risk score, and payment files (MAO-004, MOR, MMR), professional, institutional, and pharmacy claims, EHR encounter extracts (eClinicalWorks format today), lab results, vitals, and financial statements, delivered through secure cloud storage folders on a schedule the product tracks, or by manual upload.
How Alara Healthcare uses it
Onboarding is a configuration change, not a coding project: Alara Healthcare's data service sets up sources, file types, column mappings, locations, and roles, and the first screens fill after the first full delivery.
Data & EMR Integration servicesQuestions people actually ask
Data Management questions
I sent a file this morning and none of the numbers have moved. Is something broken?
Probably not yet. Arrival, reading, and correctness are three separate steps that fail independently, and each has its own screen. The Inbox shows the file was delivered. Pipeline shows what was actually read. Quality shows whether the contents hold up.
What happens when a payer changes a file layout?
The pipeline stops and asks. A schema alert names the added, renamed, or dropped column, an administrator confirms the new mapping in the two-panel editor with sample values, and the file type is versioned so the old and new layouts are both understood.
The completion share on Pipeline has been stuck at the same number for hours.
Early in a run a low share just means work is outstanding. A share that has stopped moving means the remainder is failing rather than waiting, and Batch Errors on the same screen shows what. A complete row means the row was usable, not that its contents are correct; Quality answers that separately.
See Data Management on your own file layout.
Bring a sample payer file to the demo and we will show you where it lands, what it surfaces, and who works it.