Large Language Models (LLMs) are text-generating systems that can simulate human reasoning, which opens up the opportunity to use them to handle more complex tasks than what traditional software typically manages. While traditional software might follow a pre-set sequence of instructions, an autonomous agent like Locusive has the ability to make decisions dynamically based on the context of the user’s request. Locusive can reason about what needs to be done, select the right tools, and execute steps in an adaptive, goal-oriented manner.

Locusive isn’t just answering questions or pulling data—it’s actively thinking through how best to fulfill the user's request, considering the most relevant data sources and workflows, and determining the most efficient path to an answer or action.

The Reasoning Process

When a user submits a request to Locusive—whether to fetch data, analyze information, or take an action within your product—the platform follows a structured reasoning process:

1. Determine The Action To Take:
   When Locusive receives a request from a user, it needs to determine what action to take to make progress in answering the user's question. An action can be as simple as searching for data or as complex as creating a graph on a dataset that has already been queried. At any given point in time, the system is presented with a series of tools that it can use to perform an action, and after thinking through what tool is best suited for the current task in the larger process of responding to the user, it selects the tool.

2. Running A Tool
   Once the system has decided what tool to use, it needs to run the tool. This can also be simple — searching for something on Google, or complex — creating and executing a SQL query on your database. Locusive will decide how to best use the tool, run the tool, and then retrieve the results of running the tool.

3. Evaluate Progress
   After Locusive has retrieved the results from running the tool, it decides if it has enough information to respond to the user or if it has taken the steps necessary to complete any actions necessary to fulfill the user's request. If so, it responds back to the user, if not, it jumps back to step 1 to start the whole process again. This series of steps continues until Locusive has either successfully handled the user's request, times out after 5 minutes, or determines it's unable to handle the user's request.

How To Get The Best Results Out Of Locusive

To get the most out of Locusive’s autonomous reasoning capabilities, it helps to consider a few best practices:

1. Be Clear and Specific:
   While Locusive is great at interpreting user intent, providing clear and specific instructions will lead to more accurate results. Instead of a vague request like "Show me last month's sales," try something more specific: "Show me the total sales from the US region in March 2024, broken down by product category."

2. Leverage Example Queries And Reference Guides For Database Access:
   Providing Locusive with references and example queries can help it better understand your needs. These examples give it a framework for how to structure future queries, making the reasoning process faster and more precise.

3. Be Explicit About Any Tools You'd Like To Use
   While Locusive can reason its way through a request, selecting the right tool at the right time, providing explicit instructions about how the system should handle its task can result in faster and more accurate responses.

Conclusion

Locusive is more than just an AI assistant—it’s a powerful reasoning agent designed to automate complex workflows and assist users in finding and processing data, as well as taking actions within your product. By understanding its reasoning process and applying best practices, you can maximize the efficiency and effectiveness of your team and customer interactions with your product.