Identify and select the data and AI systems

- Now that you have an idea of your company's needs, let's go deeper into how to select the right type of AI for your business. There are many different types of AI systems and tools, and choosing the right one is crucial for success. For additional context, check out the course notes for the core and generative AI landscape map. But let's start with data requirements. When you evaluate types of AI systems, consider the following. What formats of data do you need for your project? What would be your input data? What would be your desired output data? Different types of AI models and applications accept and generate different formats of data, such as text or images. How often will your data need to be refreshed? Depending on your business objective, data might need to be refreshed daily or weekly or monthly. And for some applications such as medical diagnostics, a continuous stream of realtime data might even be needed. Next, what source of high-quality data do you have at scale? It could be owned by you or coming through partnerships. Remember that the focus should be on proprietary data. For example, if your company is looking to use AI for pharmaceutical research or drug discovery, then data on molecular structures and biological pathways might be needed. So data will need to be refreshed frequently as new findings emerge. And the AI system you choose should be able to handle the scale, the complex interactions, and analyze vast quantities of data. Or if your company's planning to use AI to improve the accuracy of demand forecasting, then data on past sales performance, economic indicators, reports on industry trends and customer behavior might be needed. The system will need to assimilate large data sets in different formats and continuously update as new data becomes available and then generate accurate predictions for different products and markets. So now let's transition to the type of AI you need. This is basically the AI engine you want to implement. If your goal is to use AI to create product designs, you might want to use a system that has specific capabilities for 3D modeling or animations, or if you're interested in using AI to improve customer service, you might want a system that excels in natural language processing or chat bot development. Or, you might want a system design or product schema that is a mix of workflows and generation and automation. Once you have an idea of the type of system you need, there are three more considerations for you. The first one is implementation. How easy or difficult will it be to implement this type of AI in your company? Consider the technical resources, the personnel, the cost, and the time needed to integrate the model into your business processes. Second, scalability, can the model scale with your business? You want to choose a type of AI that will be able to grow with your company rather than becoming obsolete or struggling to handle the increased volume. And lastly, transparency. Are you able to understand how the model is making its decisions? Transparency is crucial for establishing trust, and can help you fine tune the model over time. With this information fresh in your mind, I want you to pause this video and write down your specific requirements to ensure you're selecting the best solution for your needs.