NeuReality names Shalini Agarwal to boost NR-NEXUS platform

NeuReality has appointed former Google product leader Shalini Agarwal as Senior Strategic Advisor as it expands its NR-NEXUS software platform.

Agarwal joins the Israeli AI infrastructure company after more than 16 years working on AI-led products, including senior product roles at Google Cloud and Google Workspace. At Google, she worked on Gemini integrations across Gmail, Docs, Slides, and Sheets. Earlier, she led product work for a conversational AI venture inside Google's Area 120 incubator that was later folded into Google Cloud AI.

Her appointment comes as AI suppliers and users increasingly shift their focus from model development to the operational challenges of deployment. Companies running AI systems often deal with multiple models, APIs, and computing environments at once, creating management issues around orchestration, utilisation, and cost.

NeuReality positions NR-NEXUS as a common software layer for AI inference workloads across mixed infrastructure. The platform is hardware-agnostic, supporting CPUs, GPUs, and networking infrastructure, and is designed to help organisations run and route workloads across both open-source and proprietary models and APIs.

The company is making that case at a time when many businesses are trying to avoid dependence on a single model provider or infrastructure stack. For customers, that creates practical questions about how to allocate workloads, maintain service levels, and control the cost of delivering AI services in production.

Founded in 2019, NeuReality has offices in Israel, Poland, and the United States. It has focused on inference infrastructure rather than training systems. Its broader product line includes networking and processor products for AI deployments, while NR-NEXUS sits above that infrastructure as a software layer.

Shift in focus

NeuReality argues that the industry bottleneck is changing. As foundation models become more widely available, businesses are placing greater emphasis on operating AI systems reliably and economically across different environments.

That view is reflected in Agarwal's remit. She will work with NeuReality's leadership team on product strategy as the company expands NR-NEXUS among organisations deploying AI at scale.

"AI builders should be focused on creating intelligence and user value, not stitching together fragmented infrastructure across models and environments," said Moshe Tanach, CEO of NeuReality. "As the foundation layer continues to commoditize, the challenge shifts to operating AI systems in production. NR-NEXUS is designed to be that operating layer - enabling organizations to run AI services efficiently, economically, and at scale, without needing to become infrastructure experts."

The hire also brings in application-layer experience that could help NeuReality shape its offering for customers that need AI systems integrated into business software rather than deployed as standalone technical projects.

Agarwal began her career at the MIT Artificial Intelligence Lab, where she worked on large-scale AI systems. At Google, her work spanned enterprise software and cloud services, giving her experience across products used by both large corporate customers and mass-market users.

Operating layer

NeuReality is positioning NR-NEXUS to sit between models and production applications. In practice, that means handling deployment and orchestration across varied environments while giving customers flexibility in their choice of models and APIs.

The software is intended to improve utilisation and lower the cost of AI service delivery by managing end-to-end data flows, token economics, and service-level objectives. Those issues have grown more important as enterprises move from pilot projects to broader rollouts with stricter performance and budget requirements.

For infrastructure vendors, this has become a competitive battleground. Cloud providers, chipmakers, and software firms are all trying to define the management layer that controls AI workloads in production. NeuReality is seeking to differentiate itself by focusing on interoperability across heterogeneous environments rather than tying users to a single stack.

Agarwal framed the opportunity in similar terms. "Enterprise AI is entering a new phase," she said. "The opportunity is no longer only about building better models, but about making AI systems practical to operate in the real world across different environments, workloads, and business requirements. That requires a new operating layer between models and production applications. NR-NEXUS is addressing this gap, and I'm excited to help bring it to a broader set of customers and partners."

Tanach said her background in widely used AI and software products would help shape NR-NEXUS. "Shalini has spent her career bringing AI and software products to millions of users," he said. "Her experience at the application layer will be instrumental in building NR-NEXUS to serve real production AI applications."