Demo: No-code AI with machine learning

- Now that we have practiced with other people's business cases, go back to your notepads. What kind of data have you listed? Is it numbers, strings, images, videos, or text? What kind of data you have is your first question. The second question, is your industry vertical? And the third question, is your functional area where you work? Now, you can easily pick a no-code AI that works for your data, your industry, for your business function. Let's practice with a different no-code AI tool. A business manager at a telco business wants to reduce customer churn and build an AI to predict whether a customer would churn or not, based on the customer information. Let's ask the three questions. What is the kind of data you have? Is it numbers and strings, images and videos, or language phrases? See, it's a table of numbers and strings. The customer ID is a string, gender is a string. Whether the customer is a senior citizen is a number. It's a one or zero. The demographics of the customer are all strings. The tenure of how long the customer has been at the company is a number. What services the customer has bought from the company are also strings. We are interested in paperless billing. Whether it is a yes or no, it's a string. And the last column is customer churn, again, a yes or no. This dataset has historic data of when the customer churned, which is a yes, and when the customer did not churn, which is a no. The second question is, what is the industry? It is a telecom company. We don't see a no-code AI tool for this industry. The third question is, what is the functional area of work? It is marketing. So we look for a no-code AI tool that works with machine learning. We can pick the no-code AI tool Akkio. You can enter the dataset as training data and click Predict and get the AI built out within minutes. Here is a publicly available dataset for customer churn. It's available on Kaggle. You can download and try out Akkio for yourself. Let me show you. Here, I've uploaded the telco dataset in Akkio, so I save it. Remember, you always have to start with a business problem. In this case, we are checking for customer churn and want to see whether paperless billing works to reduce customer churn or not. Remember, you are building the AI using no-code AI for problems for which you know the answer. In this case, you know if you give paperless billing, it will impact churn. You're going to make the AI do a prediction for you with your customer dataset. And I select paperless billing and churn. And the AI is built with no-code AI, absolutely with no code.