LIVINGSTON, NJ, Sep 29, 2026 – Harell Data will deploy its platform on CoreWeave Cloud, bringing AI models to proprietary scientific datasets for training, fine-tuning and inference rather than moving the data to the models. Training jobs will run on NVIDIA A100 and NVIDIA Hopper GPUs, while raw data stays inside Harell.
“High-quality, domain-specific data is absolutely a primary catalyst for the next generation of AI breakthroughs,” said Harlan Robins, PhD, founder and president of Harell Data. “We created Harell Data to connect researchers to proprietary scientific datasets for model training and inference without requiring them to view or extract the underlying data.”
“CoreWeave gives us the scale and the technical engagement to move at the pace our customers expect,” added Saray Covey, head of operations, Harell Data.
“Biotech research teams need a cloud that lets them focus on the science and discovery rather than the systems underneath it,” said Jon Jones, chief revenue, CoreWeave. “We’re partnering with Harell Data to give researchers exactly that as their platform scales.”
Model Validation
Harell’s data partners keep validation test questions hidden from model builders. The company says builders have not seen the answers, and all results are published against a common standard.
Compute Billing and Model Sales
The platform meters compute use for each run. Data owners receive a revenue share from every training job that uses their dataset.
Model builders receive the trained model, intellectual property and the right to sell it. They can list models on the Harell marketplace under a “Models as a Service” arrangement and collect a fee each time someone runs a model.
Infrastructure Benchmarks
The announcement also cited MLPerf results and CoreWeave’s Platinum rankings in SemiAnalysis ClusterMAX 1.0 and 2.0, where it said it was the only AI cloud to rank Platinum in both. Artificial Analysis ranked Moonshot AI’s Kimi K2.6 No. 1 for inference speed and price-performance.
Source: CoreWeave
About CoreWeave
CoreWeave provides cloud infrastructure and managed services for artificial intelligence and high-performance computing workloads. The company operates GPU-based infrastructure, data centers, and software for training, inference, and large-scale AI deployment. Its services include cloud computing, Kubernetes management, storage, networking, and infrastructure orchestration. CoreWeave also develops software to manage GPU resources and computing infrastructure. Its customers include AI companies, cloud service providers, enterprises, research organizations, financial services firms, media companies, and software developers. Organizations use its infrastructure for AI model training, inference, visual effects, scientific computing, and other compute-intensive workloads. CoreWeave was founded in 2017 and is headquartered in Livingston, NJ. CoreWeave reported 2,100 employees worldwide. The company operates data centers in the United States and Europe.
About Harell Data
Harell Data, founded in 2025, is a scientific computing company based in Bellevue, Washington, U.S. Its Harell Cloud service lets research teams train and run AI models on proprietary scientific datasets. Users can search datasets, provision NVIDIA A100, H100 and H200 GPUs, train models and run predictions. Datasets cover antibody-antigen binding, T-cell receptor binding, protein stability, immune sequencing and protein interactions. Researchers can track jobs through live metrics and logs, then deploy models for inference. The service runs models against data in place, keeping raw datasets in owner-controlled environments. Data providers receive revenue shares when developers train on their datasets. Model developers retain intellectual property and may earn revenue when others use their models. Data partners include A-Alpha Bio and Adaptive Biotechnologies. Harell may add materials science and imaging datasets.
