How a leading cancer center gave researchers self-service access to data that was previously locked in silos, cutting weeks of manual work down to minutes.
The UVA Comprehensive Cancer Center sits at the intersection of patient care and translational research. Each year, the center sees approximately 5,000 new cancer diagnoses and collects around 8,000 specimens, many with accompanying genetic and biomarker analysis. On paper, this is a substantial research asset. In practice, it was of limited use because the data was fragmented across systems that don't speak to each other.
Clinical information lived in the cancer registry and the hospital's electronic health record system. Specimen data was housed in a separate biorepository database. Genomics data was split across a sequencing program and a high-performance computing cluster. For a researcher trying to assemble a cohort that crossed any of those boundaries, the path forward meant contacting multiple people and waiting, sometimes for weeks.
A five-year partnership with Manifold changed that.
The challenge wasn't a lack of data. It was a lack of infrastructure to make that data findable, queryable, and actionable.
When researchers needed prostate cancer tissue samples annotated with cancer stage and demographic information, the request touched five different individuals across multiple systems and took approximately six weeks to fulfill. This was not an unusual case.
For investigators trying to move quickly on a hypothesis, that kind of turnaround is often fatal to a project. Researchers learn not to bother to ask, or they design studies around what's easy to access rather than what's scientifically optimal. Every manual data request pulls staff time away from higher-value work. External partnerships become hard to operationalize. Faculty recruitment suffers.
UVACCC selected Manifold as its research data platform, deploying a modern AI-driven environment for cancer data management across clinical, genomic, and specimen data.
Unified data, queryable in one place. Clinical, specimen, and genomics data (previously siloed across separate systems) were brought into a single, queryable environment. For the first time, a researcher could ask a question spanning patient demographics, cancer stage, specimen availability, and genomic data without assembling the pieces themselves.
Self-service cohort building, with AI. Manifold's AI-powered Cohort Explorer allows investigators to describe a patient population in plain scientific language and get results in minutes. The prostate cancer query that once required six weeks and five staff members now takes approximately five minutes with no additional staff involvement.
Advanced analytics on researchers' own data. Researchers and Translational Research Teams can bring their own data to the platform, creating a secure central repository and unlocking AI-native analysis. Work that previously required weeks and a data scientist collaborator can now be done in minutes.
Tissue sample access. Manifold surfaces specimen availability directly in the research environment, allowing investigators to identify and request relevant samples without routing through multiple staff.
As new patients are seen, specimens collected, and studies launched, UVACCC's research data asset grows continuously. Manifold's platform scales with that growth, adding new data modalities and onboarding new research programs without rebuilding the foundation.
UVACCC's partnership with Manifold is poised for continued growth. In 2026, the Cancer Center is expanding the breadth of data on the platform with deeper clinical data and genomics coverage, and more investigators bringing their own datasets. Leaders at other institutes across UVA have also taken notice and are actively exploring what a similar approach could mean for their own programs.
Forward-thinking academic medical centers and comprehensive cancer centers are rethinking what’s possible with AI-powered research and discovery.
.svg)
Platform walkthrough tailored to your data environment
Current workflow review and fit assessment
Implementation timeline and resource requirements
