Voio Exits Stealth With $8.6 Million Seed Funding to Unify Radiology AI Workflows

Voio, a specialized artificial intelligence laboratory focused on healthcare applications, has publicly launched after securing $8.6 million in seed funding led by Laude Ventures and The House Fund.

The company is developing a consolidated reading platform intended to streamline diagnostic workflows for radiologists across all imaging modalities.

This venture spins out of established research collaborations between the University of California, Berkeley, and the University of California, San Francisco (UCSF). The founding team, which includes UC Berkeley Professor Trevor Darrell, UC Berkeley/UCSF Assistant Professor Adam Yala, and UCSF practicing radiologist Dr. Maggie Chung, simultaneously released Pillar-0. This initial offering is an open-source AI model designed to interpret medical images, recognizing hundreds of conditions within CT and MRI scans with demonstrably high precision.

Pillar-0 establishes a new performance benchmark, showing a 10% to 17% accuracy improvement over current leading models released by major technology firms including Google, Microsoft, and Alibaba on standardized tests. Specifically, the model achieved an .87 AUC across over 350 distinct findings in chest CT, abdomen CT, brain CT, and breast MRI studies. This development directly addresses the increasing strain on radiology departments globally, where high annual scan volumes exacerbate workforce shortages and contribute to diagnostic delays and professional burnout. Current clinical practice forces radiologists into constant context-switching between image viewers, reporting software, and electronic health records, reducing time spent on core interpretation tasks.

Voio’s proposed platform seeks to eliminate this fragmentation by using frontier vision-language models to interpret entire exams and automatically draft comprehensive, high-quality reports for radiologist review.

Consequently, the system aims to optimize the balance between diagnostic speed and clinical rigor, allowing practitioners to focus resources on complex decision-making. The team’s prior research already holds clinical relevance, with existing AI tools having been validated across more than 90 hospitals in 30 different nations. Furthermore, their breast cancer risk prediction model, Mirai, has already been utilized in processing over two million mammograms worldwide.

Additionally, the open-source nature of Pillar-0 facilitates external validation and academic extension, promoting transparency in an AI sector often criticized for proprietary performance claims.

Voio intends to foster independent benchmarking to solidify evidence-based standards for clinical AI adoption moving forward.

The investment and platform development signal an intent to pivot radiology from a primarily reactive documentation function toward proactive, predictive medicine, identifying future health risks well before symptomatic presentation.

This foundational work positions Voio to potentially integrate across multiple specialties as their multi-modal agentic workflows mature.