ConcertAI, the leading oncology real-world evidence data and AI SaaS technology company, has officially announced a collaboration with NVIDIA to conceive a broad set of translational and clinical development solutions within its proprietary ConcertAI CARA AI platform. Leveraging NVIDIA Inference microservices (NIM), including the recently published Llama 3 NIM, NVIDIA CUDA-X microservices, and the NVIDIA NeMo platform, ConcertAI plans on delivering a leading brand of intelligence and insight across oncology research and treatment. But what does that mean on a more granular level? Well, we begin from how the stated development will bring to the fore high-performance AI models for clinical development solutions. Here, the idea is to apply NVIDIA NIMs and provide scalable, high-performance AI model deployment with low latency. This, in turn, should enhance the flexibility, interoperability, and cost efficiency of the ConcertAI CARA AI platform. The integration in question can be further expected to support clinical trial patient matching, protocol automation, and research site co-pilots through real-time analytics and model management for large-scale AI applications. Next up, we must get into the large-scale processing capabilities of multi-model data. These capabilities basically translate to harnessing of NVIDIA CUDA-X microservices to accelerate ConcertAI’s large-scale data processing pipelines. You see, by expediting data processing speeds and enabling more effective management of its vast oncology data, ConcertAI will put-together the largest curated oncology data set globally.
“Life sciences research and precision medicine both involve complex decisions based on many types of data and time points,” said Jeff Elton, PhD, CEO of ConcertAI. “For any single patient, we may have billions of unique data points, and we may be looking across millions of records, making this a domain where AI can enable insights not possible with prior technologies or methodologies. We are thrilled to work with NVIDIA to push the boundaries of what AI can achieve in oncology translational research, clinical development, and care.”
Moving on, another focal point for this integration is going to be the development of precision oncology and medical large language models (LLMs). Once developed, these models will be trained on ConcertAI’s industry-leading multi-modal data to facilitate advanced translational simulations that can then guide novel, first-in-human studies, clinical trial simulations, and design optimizations. Hold on, there is more, considering such a setup can also come in handy when the agenda is to achieve clinical decision augmentation support for clinical pathways, as well as to identify beneficial diagnostic and treatment approaches.
The entire runner provides a rather interesting follow-up to ConcertAI’s efforts over the last two years, a timeframe during which it assembled the deepest and largest multi-modal oncology data repository in the industry, representing more than 8 million patients. Building upon that, the company took over the American Society of Clinical Oncology (ASCO) CancerLinQ program back in December 2023, under a long-term cooperation agreement with ASCO. In collaboration with molecular diagnostic partners and its radiological imaging business, TeraRecon, ConcertAI has also established a research capability that spans genomic, transcriptomic, digital pathology, digital radiology, clinical, and social determinants of health data. This dataset is markedly made to cover all 50 states, and therefore, offers the broadest representativeness and generalizability of any clinical data source.
“AI offers incredible potential to transform how medicines are designed and developed, and bringing generative AI tools to improve clinical trials is a groundbreaking and necessary step,” said Kimberly Powell, VP of Healthcare at NVIDIA. “Integrating NVIDIA’s NIM’s into ConcertAI’s SaaS and extensive multi-modal data platform will revolutionize clinical trial design and outcome prediction.”