Persistent Systems has begun working with Nvidia on AI-based software for healthcare and life sciences, initially targeting computational drug discovery and virtual screening.
The effort centres on a new Persistent product, Generative Molecules and Virtual Screening (GenMolVS). Persistent plans to use elements of Nvidia's AI software stack and accelerated computing systems to develop and deploy the offering for research and development teams in biopharma and related organisations.
Healthcare and life sciences firms face growing pressure to improve research productivity while operating under strict regulation and heavy data demands. Computational methods are playing a larger role in early-stage discovery as organisations assess more compounds and reduce the number of candidates that move into costly laboratory work.
Drug discovery focus
The collaboration covers specialised research use cases, including preclinical research. Persistent and Nvidia plan to apply AI to molecular simulation and virtual screening, modelling biological and chemical behaviour before work moves into physical laboratory settings.
GenMolVS is built on Nvidia BioNeMo and the Nvidia NeMo Agent Toolkit, according to Persistent. It uses domain-specific models for molecular simulation and software agents to manage parts of the discovery workflow.
The workflow spans virtual screening, candidate prioritisation, and downstream experimental planning. Persistent said this helps teams connect simulation results to wet-lab experiments and make near real-time decisions.
Agentic systems
Persistent plans to build and deploy "production-grade agentic AI applications" for computational drug discovery using Nvidia NeMo. In this context, agentic systems refer to software that carries out tasks across multiple steps in a process, rather than generating outputs in isolation.
The company also plans to use Nvidia Nemotron open models as it develops the next version of GenMolVS. Deployment will use Nvidia AI Enterprise, Nvidia NIM microservices, and Nvidia accelerated compute and servers.
These components sit within Nvidia's enterprise software and infrastructure portfolio for running AI workloads. In life sciences, key technical challenges include handling large datasets, keeping systems auditable, and integrating tools into existing research environments.
Deployment plans
Persistent framed the collaboration as a move from experimentation to operational use in "mission-critical environments". It said the design supports simulation and inference at scale for research programmes operating under regulatory requirements.
Persistent also plans to expand its AI and large language model engineering work through Nvidia infrastructure, training resources, and certification programmes, feeding into broader data and AI platform work delivered to clients.
Ganesh Nathella, Executive Vice President and General Manager, HLS Business at Persistent, linked the product to faster research cycles and growing use of AI in discovery.
"Healthcare and Life Sciences organizations need to discover new therapies faster, but traditional R&D is too slow and labor-intensive. By combining our GenMolVS solution with NVIDIA full-stack AI platform, we enable BioPharma clients to use generative molecules and virtual screening in production so they can move from months-long experiments to AI-driven discovery in days, using simulation-led intelligence to guide real-world experimentation without compromising on scientific rigor or compliance," said Ganesh Nathella, Executive Vice President and General Manager - HLS Business, Persistent.
Nvidia's enterprise software group has expanded into industry-specific AI use cases, including healthcare and life sciences, as drug developers and research organisations look to increase throughput in discovery and preclinical work.
John Fanelli, Vice President, Enterprise Software at Nvidia, described the shift towards AI-driven research and positioned the Persistent work as part of that trend.
"To meet the urgent global demand for new therapies, the healthcare and life sciences industry is rapidly moving toward AI-driven computational research and discovery. By leveraging the full-stack NVIDIA AI platform, Persistent is empowering biopharma companies with production-grade agentic systems for molecular simulation and virtual screening," said Fanelli.
Persistent plans further enhancements to GenMolVS using Nvidia's open models and deployment tools as it rolls out agentic AI applications for computational drug discovery.