JobCare · for team discussion
These are business and policy choices — not technical ones — that the team needs to settle before we build the matching, billing, and communication rules in detail. Each is written in plain language with a few options to react to. Bring them to a working session; the answers become the spec. Items marked decide first carry the most money or legal weight.
How much risk we take on when a placement doesn't stick.
If an aide leaves soon after starting, we give the agency a free replacement (never a cash refund). This is our single biggest financial exposure — too long and we lose money, too short and agencies won't trust us.
Nurse / specialty placements cost the agency more; a longer (e.g. 90-day) guarantee could justify the premium price.
A post-op case ending on schedule isn't a failed placement — but our guarantee logic could mistake it for one.
The logic that decides who surfaces first and who we reach out to.
We score aides on reliability, skill, professionalism, would-hire-again, retention, plus role / area / language fit. The weighting decides who shows up first. We have a starting recipe — confirm it or adjust the emphasis.
We plan tiers like Gold / Standard / Watch, and a "do-not-refer" state. Decide the exact thresholds, and whether (and how) someone flagged do-not-refer can ever be reinstated.
Which complaints or ratings should hold an aide from being referred until a person reviews it.
Beyond "best match," coordinators may want to sort candidates other ways.
A stated turnaround (e.g. "candidates within 15 minutes") sets agency expectations and drives how we staff and automate.
Too many texts becomes a nuisance and a legal risk (TCPA). We need a sensible cap.
Is one opt-in to be contacted enough, or do we also need separate permission to share their vetting packet with each agency? And is a checkbox enough, or a signed form? Directly affects TCPA and privacy exposure.
e.g. available / on a job / not looking right now / can't reach. Notably: should we never contact an aide who is currently on a placement (continuity of care)?
The assistant can answer simple aide questions (pay, location, hours). What must always go to a person — anything medical? anything that sounds like a hiring decision?
Our posture is transparent, rule-based ranking with a human making the decision. Confirm we'll offer multiple sort options, keep an attorney opinion letter on file, and avoid "black-box" scoring.
Stated credentials let us start matching, but which documents must be verified before we finalize — and do we track ongoing training hours as a value-add?
There is clearly a base set every aide needs (photo ID, I-9, background check, TB test, references, résumé). Beyond that the requirements vary — we've started with role (e.g. nurses add license verification + immunizations; CNAs add a Nurse Aide Registry check). But there are likely more variables we haven't pinned down: payer/program (Medicaid vs private), care setting, patient acuity, agency-specific policy, and state/borough licensing. We need to define the full rule model so the system requests exactly the right documents per case. The app currently implements base + role as a first cut (one shared rules file).
Our data is a competitive asset, but keeping everything forever raises privacy and legal duties. We need a retention/deletion stance for profiles, messages, and history.
To avoid duplicate records when an aide arrives from a signup, a bench list, and a job site.
Today we identify a case by number only (no patient details). Integrating an agency's patient roster would sharpen matching but brings HIPAA duties.
Only once an interview is scheduled (protects our fee), or earlier (more convenient for agencies, but easier to go around us).
Or when a "bench" aide is already in our system. Priority affects who sees them first.
We want the family who received the care to rate the aide even when an agency did the placing — but we may not have the family's contact.
Which details are mandatory to get started vs. can be filled in later.
Plan is one monthly statement per agency (all that month's placements rolled up), rather than a bill per placement.
An aide works with many agencies over time, so this is judged per placement, not per aide. When a coordinator opens an aide's profile they should see the full vetting package, the aide's employment history, and their own conversations. But a placement with a competing agency is sensitive — showing that agency's name, exact dates, and outcome could expose competitive information. Working default: for a placement that was with the viewing agency, that agency (and JobCare operators) see full detail; for a placement with a different agency, the viewer sees that a placement happened, with role and rough timeframe, but not that agency's identity or the outcome.