By Ridhika Basnet · AIM Media House
On May 31, 2026 , Chicago-based health technology company, Tempus AI announced the launch of its next-generation Lens platform, an agentic AI system, connected to a massive commercial multimodal data library, which is one of the world’s largest in the field, designed to accelerate oncology drug research and development.
Lens was launched as a cloud-based data and analytics platform on May 20, 2021, where it provided on-demand access to over 35 petabytes of de-identified clinical and molecular data alongside advanced AI tools to accelerate drug discovery.
"Drug development requires thousands of pivotal decisions between molecule and approval, and at its core, it is a navigation problem—most paths end in dead studies and wasted capital, which is why the industry needs a fundamentally different approach," said Ryan Fukushima, CEO of Data and Apps at Tempus.
He also called the real-world multimodal data complex, which took weeks and months of manual analysis historically, while turning them into decisions. He further added that the next generation of Lens consolidates this workflow into a single platform.
Lens that sees the future of Oncology The next-generation version of Lens combines globally large multimodal datasets, high-performance AI computing, the oncology foundation models of Tempus, validated AI agents and scientific workflows, into an integrated single platform, Tempus said.
Lens was built with the purpose to facilitate the drug development teams design better clinical trials, by targeting patient subgroups quicker and generating critical evidence swiftly. The company also said that its multi-agent platform has several specialized tools embedded.
A custom research plan can be generated through the Lens Co-scientist AI agents, centered in oncology knowledge. Users can pitch complex biological hypotheses using fluid language, receiving a targeted analysis plan they can easily fine-tune through direct collaboration with the AI agent.
After a plan is finalized, an analysis is executed by the AI agent in code against the large de-identified patient records bearing multimodal library which shortly delivers code backed results.
Continue on AIM Media House