For most of 2024, Umair was the primary caregiver for his grandmother during her cancer treatment. She had light chain nephropathy, a hematologic malignancy monitored through serum kappa light chain levels and treated with chemotherapy, steroids, and dialysis. Umair took her to oncology four or five days a week while working remotely as a software architect. Her treatment team did remarkable work, and she had an extraordinary clinical response.
What he learned in those months changed how the founding team thought about cancer care.
Modern oncology has made stunning advances: targeted therapies, genomic profiling, biomarker-driven treatment selection. But the information that drives every treatment decision lives in long, vendor-specific reports that busy oncologists rarely have time to read carefully. Meaningful details get missed. Coordination between specialists runs on caregivers forwarding PDFs. The gap between what precision oncology could be and what it often is isn't a science problem. It's an infrastructure problem, and a software problem.
UNMIRI is built to close that gap. The first piece the team could actually solve was the one closest to engineering hands: turning the soup of vendor-specific genomic reports (Foundation Medicine, Tempus, Caris, Guardant, Natera) into one structured, citable, FHIR-aligned representation that any clinical software can build on. From there, three other surfaces followed: a decision-support API for the same software companies, a literature surveillance platform for biotech medical affairs teams, and a free interpretation tool for the pathologists and oncology fellows who interpret these reports daily.
This is the infrastructure Umair wishes his grandmother's oncologists had had. UNMIRI is being built so other families can have it.
— The UNMIRI Team
