Lab2Care

When a clinic is waiting for a lab report, staff need to find out whether the report is still pending or has arrived without a clear link to the original test order. Until that’s resolved, the clinic may be unable to give the patient a clear update. Lab2Care brings investigation, verification, and patient communication into one workflow. The demo starts with a fictional patient whose doctor ordered three tests. Two reports are available, and one is still pending. The AI agent, powered by Vultr Serverless Inference, checks the available lab records, identifies a possible match, and shows the evidence. If the reference is unclear, it drafts a clarification request. Clinic staff verify the match before adding the report to the case. Once all three reports are accounted for, a simulated clinician approves a prepared patient update. Lab2Care queues the approved message, verifies delivery to a test destination, and records when the fictional patient acknowledges receipt. A persistent timeline shows what happened at each step, so staff can distinguish a message waiting to be sent from one the patient has acknowledged. The prototype runs on Vultr Cloud Compute, with FastAPI handling the workflow and SQLite storing case progress and audit events. It uses fictional records and simulated notifications to demonstrate how AI can help investigate a missing report while clinic staff verify the evidence and a clinician controls what gets communicated.