Imports
March 14, 2025
Buildium integration activated after implementation start.
Bidirectional Buildium API brought online over an OAuth 2.0 token, replacing physical huddle printouts that carried an average 45-minute post-shift latency.
Service board
How the care loop is running.
A plain-language read on the six surfaces that keep AcuityRow useful inside a shift. Updated by our team, not by a status template.
Overall
Watching closely
Board reflects the current operating configuration, not an automated ping.
Known operating note
During Q4/holiday peaks, Care Signals ingestion may move to batch processing during 0600-0900 EST to reduce API latency. Family Loop drafts remain manual-review only.
Public site
Running steady
Marketing pages, signup, sign in entry.
No interruptions across the recent window.
Auth
Running steady
Operator sign in, role checks, approval gates.
Approval queue running on the usual one business day cadence.
Imports
Watching closely
Resident records, staffing, meds and incident feeds.
Buildium inbound batches completing normally. Queued retry window is 48 hours during an API outage.
Morning Row
Running steady
Daily brief generation and resident risk queue.
Briefs landing inside the usual pre-shift window.
Family delivery
Running steady
Approved family updates leaving the building.
All sends require a human approval. No backlog right now.
AI drafting
Running steady
Suggested family language and summary drafts.
Drafts holding inside review. Nothing reaches families without a person.
We do not publish numerical uptime percentages. The board reflects how the loop is actually serving operators, not a synthetic ping.
Recent notes
What we've worked through lately.
Survey Trail
Implementation, March 2025
Evidence packet creation moved off manual assembly.
Packet creation previously ran manually across three PDF generators. The Survey Trail endpoint now generates the packet, with SHA-256 batch verification for Batches 102 through 899.
Care Signals
Q1 2026 review
Signal weighting reduced false positives.
Weighting changes produced a 4.2% optimization in false-positive filtering against a pre-activation missed-signal error rate estimated at 6% via audit trails.
Seeing something different from what's posted here?
Tell our team