Context
Qare is a medical teleconsultation platform. For 8 months, I joined a team
dedicated to a specific and costly problem: the losses on reimbursement cases.
The problem: invisible leaks on reimbursements
Depending on the case, the user flows surfaced erroneous or insufficient
information. The downstream consequence: cases rejected by the State,
advanced amounts never recovered, and diffuse budget leaks, hard to isolate
because they were spread across many journeys and edge cases.
What was done
Our detection relied on two sources: user reports and log analysis. From
there, the core of the work was the analysis and understanding of each
problem: taking ownership of the business logic of reimbursement and third-party
payment (in coordination with the PM, through task grooming) to pinpoint
exactly where the information degraded. Then, on the dev side: adapting the
existing flows or creating new, more specific flows for the problematic
cases, guaranteeing complete data compliant with reimbursement requirements.
Stack & result
React, Redux and NestJS. Overall, the leaks on reimbursement cases were
reduced by roughly half, with on the order of 10 to 20% recovered on the
cases concerned (confidential amounts), a direct gain on revenue.