Pralia

Problem

A care case is not an event. It runs for months or years, across a hospital, two clinics, a pharmacy, an insurer and a care fund — none of which share a record. The family caregiver is the only continuous thread, and the job they end up doing is not care. It's administration.

The specific failure is reconstruction. Every time something arrives — a letter, a rejection, a new prescription — the caregiver has to rebuild the entire case in their head before they can act on it. What was already submitted. What was refused, and on what grounds. Which argument worked last time. What's due next week. That happens over and over, from memory and a shoebox of paper, usually by someone who also has a job.

Approach

Pralia is a voice-first AI care coordinator that holds the case so the caregiver doesn't have to.

One continuous history: documents, appointments, prescriptions, claims, submissions, decisions, deadlines, unresolved tasks — one model of the case, maintained across its whole lifetime rather than per interaction.

It understands change, not just content. The question isn't “what does this say” but “how did the case just change, and what does that make the next action?” A denial letter isn't a document to file; it's a deadline, a required argument and a decision point.

It keeps institutional memory — the evidence and reasoning that succeeded before, with this insurer, on this class of claim — and reuses it. Caregivers currently learn that the hard way, once, and then lose it.

And it does the work. Not a reminder app: Pralia identifies the next action, prepares the appeal or the application, and tracks the response and the deadline that follows.

Voice-first, because the people doing this work are doing it while driving, between shifts, or at a bedside — not at a desk with a form open.

Where it starts

Insurance appeals and applications for care benefits. Highest paperwork burden, clearest success criterion, and the place where getting it right returns money and care the family is already entitled to. From there: prescription renewals, appointment coordination, and follow-ups with providers and insurers.

Grounding

Co-founded with a physiotherapist and public health professional — clinical judgment inside the founding team, not consulted after the fact.

Six high-friction German caregiving workflows mapped in detail — hospital discharge, prior-authorisation appeals, the Entlastungsbetrag reimbursement loop, non-emergency transport, medication refills and family coordination — behind a ten-entity case model. Compliance was scoped before code: GDPR, BDSG, MDR, the EU AI Act, SGB V and XI and DiGA for the EU; HIPAA, HITECH, FTC and FDA SaMD classification for the US.

Presented at Silicon Saxony Days, Startups of Tomorrow, Dresden.

Pivoted from Elyra.