Original subtitles

- [Instructor] Agent AI automates AI

by making human decisions using the reasoning power of LLMs.

Can you imagine what you would like to automate

with agent AI in enterprise?

Pick any industry of your choice.

Write your ideas in a notepad.

Don't worry about whether you have the data

to build this agent AI.

We are exploring ideas without any limits.

To design agent AI to automate purposefully to help humans,

the starting point in imagining use cases is always

to think about why you want agent AI to do this work,

and what are the benefits to people.

Here are top three use cases in the industry,

which lend itself to using agent AI.

Let's talk about driving growth first.

How do companies drive growth?

They analyze customer data to develop insights

to see where they can generate incremental revenues,

where they can reduce cost, example customer churn,

and where they can innovate.

This innovation can be upsell of new products

to the same customers

or new innovative products

targeting new customer segments.

Now, can you imagine how you can build agent AI

to help drive growth?

Pause the video and write your ideas on a notepad.

You can use your own industry as an example.

We learned about how agent AI

can help with customer segmentation

in the marketing use case lesson.

Agent AI can be used to analyze customer data

and use the reasoning power of LLM

to do data analysis and arrive at insights.

This can be used for customer segmentation

or personalization insights.

A LLM can reason and can offer insights.

It cannot take action.

This is where tools come in.

The agent AI can call upon tools as code snippets

or full fledged tools to run, to drive action.

Then it can look at results

with a reasoning capability of LLM

to decide if the actions helped

or can make a decision to iterate the segment of customers

or the personalization recommendations.