5 days ago
Agentic AI: Driving Autonomy In Asset-Intensive Industries
Vivek Ahuja, VP-IT at rSTAR , spearheading business and IT transformation with a focus on manufacturing, energy/utilities and construction. getty
I've previously discussed the new frontier in AI: agentic AI. It's the next evolution in the AI timeline. However, is it hype, or will it be a game changer—especially for asset-intensive industries? If agentic AI is a game changer, how can companies get started using it?
Agentic AI continues the progression of AI. What began as rudimentary chatbots programmed to respond to queries based on keywords eventually evolved into GenAI. Using large language models and machine learning, GenAI platforms moved beyond the responses of chatbots to evolve and grow ("learn") from their interactions with people.
Agentic AI takes GenAI another step further by adding an autonomous dimension to the agentic agent. Depending on how the agentic AI agent is programmed, it can act autonomously within specific parameters or guardrails. In other words, it can take action on its own. For example, it can interact with multiple systems, such as email and instant messaging platforms, and send customers messages based on a predetermined schedule.
When integrated into your company's enterprise resource planning (ERP), customer relationship management (CRM) and other systems, agentic AI can assist with workflows, communications and tasks. Agentic AI Use Cases For Asset-Intensive Industries
Asset-intensive industries often have many complex workflows. A utility, for example, has many workflows related to field service, asset management, outage resolution and so on. In manufacturing, complex workflows include case management, support services, efficiency and compliance.
In a utility environment, outage resolution is one of the more complex workflows. It touches everything from SCADA systems and customer complaints to field dispatch, asset lookup and compliance. Today, much of this is still manual, disconnected and dependent on individual know-how.
Agentic AI acts like an intelligent coordinator, detecting issues early, planning and dispatching the right crews based on real-time availability, pulling up asset history from different systems and even guiding field techs through mobile assistants. It also keeps customers updated automatically and ensures compliance checks are in place. The real value is in how it brings everything together, reduces response time and creates a smoother, more efficient experience for the utility and the customer.
Before you rush to add agentic AI to your technology stack, consider your company's overall technical maturity. The launch of successful agents depends heavily on system integration. Agentic AI can provide insights, but it cannot take meaningful workflow actions until it integrates with core systems.
As with all AI projects, ensuring the system has plentiful, clean data is necessary. Governance and standards must be in place to ensure boundaries around AI use. Steps To Get Started
If you are interested in exploring agentic AI, here are some best practices to get started.
1. Choose a small pilot project to get started. Focus on projects with clear, tangible metrics and ROI to assess the benefits.
2. Focus on a narrow use case to reduce the risk of scope creep. It helps focus the project on specific, measurable uses.
3. Ensure systems are integrated before adopting agentic AI.
4. Follow a replicable pattern for the project, beginning with a clear, written proof of concept, a pilot project and then the production of the agentic AI. Include measurable results to assess project success.
5. Build governance into the project. Don't tack it on as an afterthought. Too many companies build AI models and try to apply governance later, which runs the risk of project delays and problems. Agentic AI: Your Personal Assistant
Agentic AI won't replace workers. Instead, it extends their abilities through autonomous, guided actions. It offers efficiency in many industries but can be especially advantageous for asset-intensive industries such as energy and utilities and manufacturing.
To achieve AI project goals, companies must ensure they have the right building blocks in place: clean data, an identifiable and narrow use case, system integration and governance built into the project from its inception. With the right items in place, agentic AI can be a powerful force for change.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
