SAP Conversational AI (CAI) is SAPs end-to-end platform for building enterprise-ready bots. This platform enables developers to build and monitor digital assistants with the intent to simplify interaction with software and raise customer experience.
SAP Conversational AI is part of the PaaS offering SAP Cloud Platform. This means the chatbots run in a separate cloud environment to keep the impact on existing processes low. Bots can integrate with other services available on SAP Cloud Platform like Document management, Data management, Analytics, or customer-developed services.
Short Introduction: AI vs. ML vs. DL
Artificial Intelligence (AI) is the academic discipline of getting computers to perform tasks that are regarded as uniquely human. The term itself does not say anything about how problems are solved.
Machine Learning (ML) is a subset of AI. It is not focused on the output of the computer (doing something intelligent) but replicating how humans learn new skills. The base idea is to feed an algorithm a lot of data and let the computer figure things out. In most cases, ML algorithms require structured/labeled data. If you think about how you’ve learned to read the process is quite similar. In Machine Learning, the system learns based on the provided data.
Deep Learning (DL) is a subset of machine learning that enables computers to solve more complex problems. This is usually done by nested layers in neural networks. Data is transformed through these layers, the data is abstracted as it is passed deeper into the network. In image recognition, the first layer could abstract the pixels of an image. The next layer encodes edges, the next one arrangements of edges, and the final layer detects a mouth or an eye.
Enterprise Use Cases
There are multiple use cases where chatbots can deliver value to businesses by reducing support expenses and response times.
Customer Support: Chatbots can handle first-level support in real-time and around-the-clock. Bots are capable of answering repeating questions regarding order status or payment issues. They can also help to gather further information to ensure all required information is available for the second support or to create support tickets.
Vendor Support: For example, bots can provide information regarding payment status. The vendor could request a list of the latest invoices and get information about the payment status, the date the payment is scheduled, or if any issues lead to a payment block. A chatbot could also help to resolve the payment block issue in case it is caused due to missing payment information.
Human Resources: Provide answers to recurring questions like “What is my available leave balance?”, “How many sick leave have I used?”. Other use cases would be booking annual leaves or change of employee master data.