AI Solution Tutorial

AI Solution Development and Registration

In Mellerikat, you can create AI Solutions to solve domain-specific problems using the ALO framework and AI Contents.

A Linux development environment is required to utilize ALO. In the ALO framework, parameters for AI Contents are adjusted to fit domain data, and experiments are conducted to create problem-solving AI Solutions.

To easily experience this process, a Google Colab notebook is provided.

Create an AI Solution following the tutorials for AI Contents.

• Create an AI Solution to detect churn in credit card customers based on TCR. (TCR Tutorial Link)

• Create an AI Solution to detect anomalies in lunch boxes based on VAD. (VAD Tutorial Link)

For detailed information on developing AI Solutions, refer to the ALO Manual.

---

AI Solution Operation

AI Conductor

Once the AI Solution is registered, it can be confirmed on AI Conductor. AI Conductor allows you to create and operate the models of the AI Solution through Edge Conductor.

For detailed features of AI Conductor, refer to the AI Conductor Manual.

The registered AI Solution is prepared for operation by creating a new Stream on Edge Conductor.

Edge Conductor

By creating a Stream, you can set the parameters for the training and inference of the AI model and configure the pipeline actions. The trained model is deployed to the Edge App to provide the service.

For detailed features of Edge Conductor, refer to the Edge Conductor Manual.