The Allen Institute for AI (Ai2) on Wednesday launched Asta DataVoyager, a data analysis platform designed to help scientists transform complex datasets into actionable insights without requiring advanced programming expertise.
The tool addresses a persistent challenge in modern research: While scientists across disciplines have access to vast amounts of structured data, many lack the technical skills or resources to analyze it efficiently. DataVoyager allows researchers to upload datasets and pose questions in plain English, receiving statistically rigorous results complete with reproducible code and visualizations.
“[Artificial intelligence] can only accelerate science if it is as rigorous and transparent as science itself,” Ai2 CEO Ali Farhadi said in a statement. “With Asta DataVoyager, we are giving researchers a trusted partner that puts powerful analytical tools directly into their hands while preserving the standards of reproducibility and security that the scientific community depends on.”
After users upload data and submit questions, DataVoyager generates clear answers, copyable code for full reproducibility, interpretable visualizations, and a methods section documenting the analytical approach and statistical tests employed. Researchers can refine their analyses through follow-up prompts, building a traceable workflow similar to a Python notebook.
The Cancer AI Alliance, a coalition of four major cancer centers, is piloting DataVoyager to accelerate cancer research while maintaining patient privacy. The alliance is analyzing de-identified patient records that remain within their home institutions, enabling cross-institutional insights without compromising data security.
One active study examines lung cancer treatments and outcomes across centers, comparing factors such as surgical timing with neoadjuvant chemo-immunotherapy, the impact of adding immunotherapy following definitive radiation, and the effectiveness of targeted drugs versus platinum-based chemotherapy.
“When I think about the future of where I want it to go, I think about this tool in the hands of clinicians, helping to answer important questions that will ensure the best possible care for cancer patients,” said Jeff Leek, chief data officer at the Fred Hutchinson Cancer Research Center.
Clinicians at the Paul G. Allen Research Center at Swedish Cancer Institute are also testing the platform. With substantial volumes of structured health data but limited data science capacity, the institute views DataVoyager as a means to empower physicians with direct analytical capabilities.
DataVoyager was built specifically for sensitive research environments. Organizations maintain complete control over their data through flexible deployment options, including hosted portals, on-premises installations, or private cloud infrastructure. Datasets can be removed at any time, allowing researchers to establish their own data governance protocols.
“Instead of asking researchers to become programmers, Asta DataVoyager lets them ask questions about their data in their own words and receive answers they can trust,” said Bodhisattwa Prasad Majumder, a research scientist at Ai2. “Our goal is to shorten the distance between a researcher’s idea and a reproducible scientific result.”
The launch represents another step in Ai2’s broader mission to develop an open, principled ecosystem for scientific artificial intelligence. As testing continues in clinical and federated settings, the institute plans to expand DataVoyager’s capabilities to support additional research domains and workflows.
