Spring Product Showcase 2026: Rescale Introduces Agentic Digital Engineering to Accelerate AI-First Product Development

Garrett VanLee

Agentic AI is already changing what the speed of product development looks like. Now most R&D organizations are still trying to close the gap between what AI and agents are capable of today and the work their teams actually do routinely everyday. Rescale’s Spring 2026 Showcase is here to solve this gap, building on a comprehensive digital engineering platform with new capabilities designed to work with the simulation workflows engineering teams already rely on.

In this blog we’ll cover the three new capability areas, now available on the Rescale Platform:

  • Agentic Digital Engineering – Simulation-native prebuilt agents automate manual tasks like configuration, troubleshooting, and report generation while keeping experts in control.
  • AI Physics – A comprehensive AI physics operating system transforms fragmented simulation data into AI-ready datasets, validated surrogate models, and production-ready inference.
  • Compute Economics – Cost and productivity controls help engineering and IT teams accelerate R&D velocity, manage cost, and optimize infrastructure usage at scale.

Agentic Digital Engineering: Agent-Accelerated Innovation

The biggest productivity drag in simulation-driven engineering organizations is the manual, repetitive work surrounding it: validating inputs before a run begins, diagnosing failures after one ends, compiling results into reports, and configuring the next iteration.

AI engineering changes that. With Rescale’s Spring 2026 Showcase, R&D teams can now deploy simulation-native agents across the product development lifecycle, with human-in-the-loop control at every key decision point.

What Agentic Engineering looks like in practice:

  • Prebuilt agents for everyday simulation tasks. Engineers can access Rescale’s growing library of purpose-built agents for input file validation, job failure diagnosis, solver-specific troubleshooting, hardware configuration and benchmarking, automated report generation and more.
  • Workflows with agent nodes. Simulation managers can encode best practices into controlled, multi-step workflows. Agent nodes run automatically in context, pass outputs to downstream steps, and keep engineers informed at defined checkpoints rather than requiring active monitoring throughout.
  • Security and governed control are built into the platform. Rescale agents operate under explicit autonomy levels, from observe-and-recommend to act within specific guardrails, allowing teams to expand automation incrementally. Execution ties directly to existing access controls and organization policies.
rescale agent library
Launch prebuilt AI agents from Rescale’s Agent Library, specialized to tasks throughout the modeling and simulation lifecycle.

Daikin Industries, one of the world’s largest and most innovative HVAC and industrial manufacturers, is building toward an AI-first R&D ecosystem on the Rescale platform. After deploying Rescale for cloud computer-aided engineering (CAE) and data intelligence across R&D sites, Daikin significantly reduced manual simulation data-management effort and is now advancing toward broader agentic digital engineering capabilities across its global R&D organization. 

“Daikin has a clear vision for what AI-driven engineering excellence looks like across our global R&D organization. What excites us about this moment is how directly Rescale’s new capabilities align with our vision for the future of industrial manufacturing. We are already seeing productivity gains today, with a roadmap of future capabilities that matches our ambition for what comes next,” said Satoru Takanezawa, Senior Engineer and Group Leader, Digital Engineering Group, Technology and Innovation Center, Daikin Industries.

McLaren Automotive, whose partnership with Rescale was announced at GTC 2026, is an early example of what this looks like at scale. Running on a platform trained exclusively on McLaren’s own engineering data and powered by NVIDIA infrastructure, McLaren’s engineers can now evaluate thousands of design iterations in hours across multiple physics domains, with agentic workflows delivering a 3x boost in expert productivity.

AI Physics: A Complete Operating System for Surrogate Modeling

Surrogate modeling has been on engineering roadmaps for years. Neural networks, reduced-order models, and machine-learning accelerators are well understood in concept, so the challenge today is not about understanding the technologies. It is now about finding a practical, and scalable method of turning raw simulation data to production-ready models that actual engineering teams use and trust.

ai physics surrogate modeling
AI surrogate models provide high-accuracy predictions of aircraft performance and structural response.

Rescale’s AI Physics operating system (OS) solves the full pipeline in one integrated environment:

  • Data to model in a single guided environment – The AI physics OS connects simulation data management, model training, versioning, and deployment without requiring engineers to stitch together separate tools or involve dedicated data science teams.
  • Production-ready model architectures – Engineers get no-code access to leading architectures including NVIDIA PhysicsNeMo and multistream transformers, enabling adoption of AI-accelerated methods without deep machine learning expertise.
  • Inference in the tools engineers already use – Surrogate model inference runs directly inside Autodesk Alias and Blender, allowing design teams to evaluate thousands of design variations in their native environment and compress iteration from days or weeks to near real-time.
  • Open, flexible architecture – Teams can use open-source or custom frameworks and embed third-party AI physics tools without vendor lock-in, and swap tools as the landscape evolves.

The data preparation bottleneck is addressed directly through a copilot that guides engineers in converting their own historical simulation data from across their CAE environment into a training-ready dataset, bridging the gap between agentic digital engineering and AI physics without custom scripting or data engineering work.

Customers deploying AI physics report measurable gains across the product development lifecycle:

  • 4x more design candidates evaluated per development cycle
  • 30x cost-efficiency improvement
  • 60% faster product development

Compute Economics: Smarter Spend, Better Performance

Cloud HPC gives engineering teams access to enormous computational scale. Without the right controls, that scale can mean cost exposure, varying throughput, and hardware selection friction that consumes engineering time better spent on most strategic tasks. As more resource and compute decisioning becomes agentic, these challenges only grow, making automated, rules-based controls and real-time visibility critical to ensuring that AI-first initiatives are also cost-effective.

Rescale’s compute economics capabilities introduce a new layer of intelligence and active control:

  • Compute Cost Controls – Unique org-wide controls that allow engineering and IT leaders to easily make tradeoffs between cost savings and availability of computing resources. Leaders can monitor utilization and make transparent, computing optimizations to meet program-level goals without manual configuration or engineering overhead.
  • Coretype Collections – Curated hardware groupings that deliver equal or better performance than single-coretype selections. The platform can automatically provision resource upgrades including improved memory, storage, or network bandwidth at no additional compute cost, removing the hardware benchmarking burden from engineering teams.

Delivering HPC in the cloud provides the scale organizations need. These capabilities add the economic control layer that makes that scale both cost-efficient and performant at the organization and project level.

Continuous Innovation Across the Platform

Beyond agentic AI and computing economics, the Spring 2026 release includes meaningful advances across the broader Rescale Platform.

Data Fabric for Connected Digital Thread extends Rescale’s data layer to capture, connect, and activate engineering knowledge across systems of record. A growing library of connectors supports platforms like Microsoft SharePoint, AWS S3, Azure Blob, and more.

unified data fabric
Unified data fabric links SharePoint and other key enterprise tools to power AI-assisted engineering decisions

Workflow Builder for Advanced Modeling and Simulation gives simulation managers a drag-and-drop canvas to encode validated processes into controlled, repeatable templates, with structured data handoffs between steps and a guided execution experience for team members running them.

These capabilities reinforce each other: Data Fabric feeds the intelligence layer that agents and Rescale AI Physics models draw from; Workflow Builder structures the processes they operate within.

Job Troubleshooting agent embedded into a predefined simulation workflow automatically diagnoses simulation issues.

Built for the AI-First Era

Rescale’s Spring 2026 release reflects how far Rescale’s platform vision has come. Organizations that can systematically capture, structure, and act on simulation knowledge will run faster, iterate more broadly, and compete more effectively. Rescale’s digital engineering platform provides a practical, production-ready path to get there, with immediate workflow improvements from day one.

To see agentic digital engineering, AI physics, and compute economics in action, visit the 2026 Spring Showcase landing page, read the announcement, request a demo, or connect with your Rescale account team.

×