FlexAI Chat Assistant

Navigating complex datasets and deriving actionable insights can be a daunting task for many users. Traditional methods often require intricate querying or an in-depth understanding of the platform’s interface, making it challenging for new users to obtain the precise information they need promptly.

FlexAI Chat Assistant is our take on a generative chatbot that empowers users to converse with their data in natural language, allowing for a seamless blending of analytics with the simplicity of everyday conversation. With the chatbot, GoodData addresses the existing barriers by delivering quick, accurate answers to business queries without the intricacies of manual analytics design.

With the FlexAI Chat Assistant, users can:

• Interact directly with their data in a conversational manner.
• Generate analytical objects like metrics and visualizations instantly.
• Obtain accurate answers to specific business queries without manual analytics design.

Ask Your Data Questions

Engage with your data by asking straightforward questions, and the chatbot will provide the answers for you.

Steps:

1. Navigate to the “Ask about analytics” search box on the Homepage.

   
   

2. Input your query, such as “How many products were sold over the past year?”.

   A chatbot window opens for your conversation. The chatbot processes your question and gives you the answer:

   
   

3. You can continue the conversation from there. The chatbot is context-aware, so you do not need to respecify it in every question. For example, I can now ask the chatbot break down its previous answer by category:

   
   

Understand Where the Answers Come From

As an added layer of transparency, you can view details of the query the chatbot used to arrive at its answer. This should help you understand where the data is coming from, and double check the chatbot is understanding your question correctly.

Steps:

1. Click the “!” icon next to a chatbot answer.

   The answer is expanded showing key details of the query used to arrive at the chatbot’s answer:

   
   

2. If you don’t like what the query is including, you can ask the chatbot to modify it further:

   
   

Create Visualizations on the Fly

Chatbot can also generate visualizations!

Steps:

1. Ask the chatbot to present its answer using one of the supported visualization types.

   The chatbot will generate the visualization directly into the conversation:

   
   

2. You can ask the chatbot to edit the visualization for your. For instance, I would like to rename the unseemly m_quanitty_sum to something more descriptive, such as Units Sold:

   
   

3. You can save the visualization for later use or further editing, and you can add it directly to a pre-existing dashboard:

   
   

Usage in the Demo Environment

To test this feature you need to integrate your demo environment with OpenAI using your own OpenAI API token, see Start Using the Demo Environment.