Automating Resource Management with AI Agents

Imagine having a system that responds to all your queries, no matter how complex, or even fills out intricate forms on your behalf to obtain services. It's especially helpful for beginners who want to retrieve data effortlessly. Now, picture setting up an information hub where your 
clients can access what they need instantly, with minimal human involvement. Just simple, intuitive interaction—no technical jargon, no complicated commands—just natural language queries that get results.

Yes, the new HCL AI agents make it a reality. By introducing workflows and tasks, we have automated processes and procedures and reduced human efforts. Now, we have reduced human efforts further with the new AI agents. A proxy or a substitute that can work on your behalf to manage the operations for getting the required information or to initiate a process. Yes, create an AI agent and relax, it gets the job done and provides you with the desired results. 
An AI agent acts as a bridge that connects the resources of HCL Universal Orchestrator with Genetic AI. The AI agent with the LLM services manages resources such as workflows and other tools, determining what to run, when to run it, and how to run it, to get the results. Most importantly, it operates autonomously, eliminating the need for human intervention. ​

If you are a service provider, with an AI agent you can set up an infrastructure or hub of information that can be accessed by your clients for information. Now, how exactly can you create this infrastructure? How do you make sure that the clients are provided with the correct information all the time and without any complex steps? 
Create workflows and tasks, or identify tools that address all possible scenarios related to your service. Yes, automate steps through workflows that generate results or retrieve data or perform actions to meet client needs. Then add all these workflows and tools in the AI agent definition. Next, connect these resources to Gen AI services for smart management. To accomplish this, we can utilize Large Language Models (LLM). You can leverage three agent types to include in the AI agent definition: 
1) Basic Agent – A simple agent type that can be customized directly within its definition. You can define its role, set specific instructions, and establish guardrails to control its operations. This type is ideal for straightforward, single-task automation. 
2) Agentic AI Builder Agent – Designed for more complex behaviors, this agent type leverages the Agentic AI Builder platform to integrate multiple LLMs and tools. It can execute intricate, multi-step processes using diverse resources, including workflows defined in the agent configuration. 
3) Custom LLM Agent – If you have your own customized LLM, you can integrate it using the external MCP option, allowing full flexibility in how the AI agent operates. 
All of these agent types can be added to the AI agent definition. 
Additionally, the AI agent allows you to select from a wide range of LLM services. Once configured, clients can interact with the agent through a simple chatbot interface—no complex commands or jargon required. Users can enter free-form text to ask questions or initiate operations, and the selected LLM will manage the associated workflows to deliver the desired results.

Let's decode this process with a real life scenario. 
Consider an insurance company that manages policy updates, renewals, cancellations, customer inquiries, or claims. The workflows and tasks that can potentially cover all the scenarios related to taking a new policy are defined and once the specific details of a client are provided, the process can be initiated. 
Now lets say, one of the clients wants to introduce a new policy. They have to provide details manually to initiate the process. Now think about a system or interface which does it for the clients,getting the necessary information from the clients and creating a new policy with no manual effort. To accomplish this goal, define an AI agent by integrating the workflows and tools, also an LLM model of choice. Now the clients can interact with the AI agent using a chatbot. The clients do not need to be technically-savvy, they can interact with the AI agent in free-form language to get the job done. Once the necessary information is provided—or if any details are missing—the AI agent will prompt the user to supply the required inputs. It then orchestrates the associated workflows to create a policy for the client, ensuring a smooth and automated process. 
By effortlessly integrating HCL Universal Orchestrator's resources with Genetic AI and Large Language Models, AI agents provide on-demand information and initiate processes with minimal human intervention. This not only enhances efficiency and reduces manual effort but also improves client satisfaction through intuitive, free-form language interactions. From streamlining complex processes to serving as a central information point, HCL AI agents represent a significant leap towards truly autonomous and responsive service delivery.

---

Cherian T M, Senior Technical Writer at HCL Software
Cherian T M is a dedicated technical writer who transforms complex software features into clear, accessible content. With expertise in creating documentation, engaging blog posts, and informational videos, he specializes in making new product capabilities easy for users to understand and adopt.  
​When not charting the latest feature release, he enjoys the active pursuits of playing cricket, finding new perspectives through solo motorcycle rides, and diving into a good book or planning a new travel itinerary.

---