Context Builder Guide
  • Context Builder Guide
  • Welcome to Context Builder
    • 💡Quickstart Guide
    • ✨New to Generative AI?
  • BUILDER GUIDE
    • 📄DataSet
      • Data Loaders
      • Embeddings
      • Vector Databases
    • ⌚TOOL
      • Conversation chain
      • Conversation retriever chain
      • Self Checking
  • APP ANALYSIS
    • Monitoring
  • User Case
    • Simple Conversation
    • Product customer service
    • Mock Interview
  • User Agreement
    • Terms of Service
    • Privacy Policy
Powered by GitBook
On this page
  • Step One
  • Step Two
  • Step Three
  • Step Four
  1. User Case

Product customer service

PreviousSimple ConversationNextMock Interview

Last updated 1 year ago

Do you often see 24-hour product customer service on the official websites of major products? These product customer service representatives promptly answer users' questions and soothe their emotions, providing crucial service guarantees.

With the context builder, you can quickly create such a product customer service robot. Let's take a look together!

Step One

Create a new dataset and upload the company's product documentation.

Please ensure that the content of the product documentation you upload is accurate because the robot will respond based on the content of the uploaded document.

Currently, only PDF format is supported. In the future, we will support various file formats such as Notion documents, Google Docs, etc. Stay tuned.

Step Two

Click the "Add Task" button on the blank workflow page and add a Tool-Conversational Retrieval QA.

Step Three

Complete the configuration for the Tool-Conversational Retrieval QA on the right side.

  1. You can change the model type, with Gpt-3.5 and Gpt4 currently available.

  2. You can adjust the model's settings, tweaking parameters to obtain different prompt outcomes, such as Temperature and Top_p.

  3. Fill in the System Prompt, for example: "You are a product customer service, please answer users' questions politely."

  4. Most importantly, don't forget to link the Data, selecting the Dataset you just created.

Step Four

Debug and publish your app. It's advised to always debug before publishing your app.

If the results aren't satisfactory, you can promptly adjust the model parameters and System Prompt.

In just four simple steps, you can create an AI chat application capable of answering questions based on private data.

During conversations, it will prioritize searching the content within the private data as background knowledge for its responses. It's particularly suitable for the following scenarios:

  1. Reading documents｜books

  2. Product customer service/employee assistant: Upload some private data of the enterprise (product manuals, administrative management systems, after-sales rules)

  3. Advanced role-playing: Upload detailed information about a character to create your unique character.

Below is an example of an electric vehicle product customer service bot, which can answer product-related questions, including shipping, discounts, invoices, and other common questions.

SAP
ORACLE