Lens replaces chance with evidence.

Four layers of agentic infrastructure, built on the world’s largest real-world multimodal dataset.

Overview

The agentic AI platform, purpose-built for drug development.

 

Fueled by the largest real-world multimodal dataset, with AI agents that design, execute, and interpret scientific workflows — expanding what your team can do.

Layer 01: Multimodal RWD

The world’s largest real-world multimodal dataset, built for AI.

 

9M+ patient cases. DNA, RNA, digital pathology, radiology imaging, and clinical outcomes linked at the patient level.

 

Lens inverts the standard framework of customer-supplied data, vendor-defined model.

 

A decade of data engineering and harmonization sits underneath Lens, giving your team immediate access to a unified, analysis-ready platform. Upload your own proprietary data or workflows when needed for co-analysis with Tempus RWD.

 

AI-ready by default. Access curated structured and unstructured records for foundation model training. Medications, diagnoses, demographics, biomarkers, surgical, and radiation histories are all extracted from unstructured clinical notes and ready for analysis.

9M+
total patient records
10M+
whole slide images
4M+
samples sequenced
~400K
DNA + RNA profiles
7K+
connected healthcare institutions

Layer 02: On-demand Compute

Zero-friction. Pre-configured. Integrated.

 

Eliminate the operational overhead of cloud architecture setup and massive data transfers.

 

Instantly spin up pre-configured, on-demand compute environments featuring 96 vCPUs, with the ability to scale to multi-GPU configurations. Train, fine-tune, and deploy models right next to your data within a secure, fully governed, and compliant environment—no engineering tickets or data pipeline transfers required.

Layer 03: AI Workflows and Models

The right model for every task.

 

Frontier large language models plus domain-trained foundation models.

 

Lens is model-agnostic, ensuring every query dynamically routes to the optimal model for the job. Frontier LLMs effortlessly handle reasoning, planning, and code generation, while Tempus’ proprietary, domain-trained multimodal foundation models bring the necessary scientific rigor.

 

Model selection is transparent. Every AI-generated insight exposes the underlying model choice, pipeline parameters, and the auditable source code — allowing technical teams to seamlessly review or expand the workflows directly in Workspaces.

Digital pathology

 

Biomarker prediction trained on millions of de-identified slides.

Multi-omic models

 

Fuses DNA, RNA, and clinical signals at the patient level.

Outcome prediction

 

Grounded in patient-level real-world outcomes data.

GPT · Claude · Gemini · Llama

 

Routed in for reasoning, language, code.

Layer 04: Intuitive Interface & Workspaces

A unified, collaborative environment designed for immediate discovery and advanced coding.

 

Access point-and-click analytical tools for instant insights. Quickly build cohorts, query clinical notes, and run complex analyses right out of the box — no code required.

 

For deeper research, migrate your work into a dedicated, secure coding environment featuring Python, R, Jupyter, and VS Code. Mount your own proprietary data directly into your Workspace for deeper analysis right alongside Tempus’ 9M+ real-world dataset.

 

Workspaces are ready to use from day one, fully pre-configured with multimodal data, notebooks, GPU-powered training and inference, and version control.

Tempus Lens chat inside a Workspaces coding environment, generating a gene co-expression analysis script

“Build a PIK3CA-mutant HR+/HER2- cohort enriched on RNA pathway activation.”

Tempus OneCohort assembled, validated, exported

“Run feasibility for a 240-patient trial in this indication.”

Tempus OneSite list, control arm, enrollment projection

“Draft the evidence pack for the IND amendment.”

Tempus OneRegulatory-grade output, audit trail attached

Security

Protected by industry-standard encryption.

 

Essential security measures for drug development.

 

Data protection wraps each layer for security and control.

HIPAA compliance

HIPAA compliant

All real-world data is de-identified under Safe Harbor or Expert Determination methodologies before reaching any model or agent.

End-to-end encryption

End-to-end encryption

All data is encrypted in transit and at rest. Keys are managed inside the customer environment.

Role-based access control

Role-based access (RBAC)

Permissions are scoped to role and program. Access is authenticated, authorized, and logged.

No AI training on customer data

Zero AI training on customer data

Customer data is never used to train Tempus foundation models without your permission. Architectural commitment, contractually enforced at the infrastructure layer.

What buyers ask first.

Replace chance with evidence.

30 minutes. Bring your program. We’ll show you what Lens can do for you.

Book a demo