The AI-watching world is enthralled with AI agents, those increasingly ubiquitous bits of autonomous software. For the uninitiated, AI agents comprise software, including a persona, reasoning, memory and operational tools to reach human-defined goals.
Once assigned goals, many AI agents execute their tasks with limited human interaction, autonomously making decisions and solving problems. They achieve their objectives by leaning on their reasoning, often based on a large language model (LLM), as well as varying degrees of short, medium and long-term memory.
Learning. Autonomy. Memory. Tool usage. Sounds a lot like skills humans possess, right?
Indeed, given AI agents’ ability to execute complex tasks, it’s clear why organizations are excited about the potential for agents to serve as virtual work collaborators, unlocking innovation.
In fact, McKinsey believes agents’ ability to automate complex business processes could help organizations push as much as 90% of high-value generative AI use cases from pilot mode into production.
AI agents fuel design innovation
As it happens, some Dell customers are already deploying agentic AI in their operations.
InstaDeep uses agentic AI to automate the design and development of printed circuit boards (PCBs), which are critical components in everything from smartphones to fridges to industrial machinery. The design tool, dubbed DeepPCB, employs a “smart agent” to autonomously optimize PCB designs, eliminating manual work while significantly reducing design time for manufacturers.
To provide DeepPCB with enough compute horsepower, InstaDeep used its Kyber supercomputing cluster. Kyber is fueled by the Dell AI Factory with NVIDIA, whose compute and networking capabilities help the company train large-scale AI models and deploy them into production.
“The Dell AI Factory with NVIDIA is the fundamental building block that allows InstaDeep to deliver its magic,” explains Karim Beguir, Co-founder and CEO of InstaDeep.
AI innovation often requires help from a robust ecosystem of partners. SCC France—a strategic partner of Dell—worked closely with InstaDeep to tailor AI and IT infrastructure to their specific needs.
In the future, InstaDeep expects to expand its use of agentic systems to physical, humanoid robots capable of executing tasks autonomously.
Agentic AI fuels sustainable solutions
Autonomous humanoid robots will wow anyone who is a fan of bleeding-edge innovation. You know what else is cool? Organizations that transform themselves, as well as the industries they serve, through AI technologies.
Meet Worley, which delivers critical infrastructure needed to meet today’s demand for energy, chemicals, and resources—while helping to shape a more sustainable and resilient future. Worley is actively using GenAI and exploring AI agents to supercharge productivity.
Eyeing more autonomous operations, Worley is looking at how to leverage agentic AI solutions to automate complex tasks, reducing manual work for its workforce.
Worley’s exploration into AI agents builds on existing GenAI projects.
To accelerate and support customer projects, Worley is creating a suite of GenAI-powered tools that will enhance workforce efficiency, improve supplier collaboration and optimize project benchmarks and estimates. The company is also eyeing AI solutions to optimize construction pathways for engineering projects, content creation and simulation. But Worley can’t execute all these AI projects alone.
To accelerate the pace of innovation, Worley partnered with Dell Technologies, NVIDIA and World Wide Technology (WWT) to design and build its own AI factory on premises, fueled by the Dell AI Factory with NVIDIA. Additionally, the Dell Professional Services team provides workshops, knowledge transfers and managed services to enable continuous optimization and operational efficiency.
The rise of agentic AI was a key factor in Worley’s choice of technology partners.
“The technology in this space is changing so rapidly. Partnering with Dell and NVIDIA keeps us ahead of that pace,” says Summer Husband, Worley’s senior vice president of data products and applied intelligence.
The future of Agentic AI
To make agents work, organizations must decide the best use cases, embrace interoperability protocols and honor cybersecurity standards—all while navigating an agentic landscape that is evolving faster than GenAI.
Soon agents may coalesce into an interconnected layer of learning systems that collaborate across vendors, domains and interfaces, according to research from MIT NANDA. In this agentic Web, dynamic agents will negotiate tasks, share context and coordinate action, effectively running and managing business operations.
Amid this shifting agentic AI landscape, organizations must make critical infrastructure decisions and choose trusted technology partners.
No matter what your environment looks like, Dell and NVIDIA can help your organization build and scale AI systems—today and in the future.
Learn more about the Dell AI Factory with NVIDIA.
