This guide will help you navigate the Hello Client chatbot settings page and provide insight on how to overcome some common challenges, enabling you to fine-tune your chatbot functionality.
Workspace Manager
Workspace is where you manage your chatbot with customized Open AI API keys, custom domains and your team email accounts.
Go to Workspace Manager, create a name for your workspace. Under your workspace, create a chatbot. Different BYOK(bring your own key) API key can be set up for each workspace. You can also invite at least one team member to manage the chatbot under each work space. One email account currently can only create one chatbot.
General Settings
Your chatbot’s settings page is its command center, a place where you can make key adjustments to your bot’s behavior and functionality. Let’s take a look at each setting and what it accomplishes.
Open AI
Base Prompt
The base prompt allows you to establish the nature of your chatbot’s interactions and give your bot a personality. You can adjust and modify the base prompt to better suit your needs. For example, you can:
- Modify the bot’s personality
- If you’d like your bot to have a casual and friendly tone, you can experiment with a phrase like this in your base prompt: “You are a friendly and casual AI Assistant.”
- Change how the bot responds to unknown queries
- Instead of saying “Hmm, I am not sure.”, you might want it to say something like, “I’m sorry, I don’t have the information you’re looking for, please contact customer support.”
- Direct its focus on certain topics
- If you want your bot to be a specialist in a certain area, you could add, “You are an AI Assistant who specializes in providing information about environmental sustainability.”
- Define its boundaries
- If you want to restrict your bot from providing certain types of information, you could specify, “Do not share financial advice or information.”
By tailoring your base prompt, you can ensure that your bot provides the necessary information, and does it in a way that aligns with your brand’s voice and values. Furthermore, it can help your bot avoid hallucinations, and prevent it from answering questions outside the scope of the data provided.
Model
This setting allows you to choose the Open AI model you prefer your chatbot to use. By default it’s set to the gpt-3.5-turbo model, which uses one credit per message. You can opt to use gpt-4, which provides more accurate responses, but it uses 20 credits per message. If you run out of credits and are on the Unlimited plan, you can use your personal Open AI API key. You can join the wait-list to attain a gpt-4 API key by sending a request to Hello Client’s support email.
Temperature
The temperature corresponds with the “creativity” of the bots responses. This value is set at zero, which instructs your chatbot to choose the most likely output when generating responses, resulting in more consistent and less random answers. You can adjust this number and experiment with the bot to fit your needs.
Access
Visibility
You can set your chatbot visibility by selecting ‘private’ or ‘public’. ‘private’ means that only you have access to the bot, and cannot embed it on a site. ‘public’ means that anyone with the link can chat with your chatbot if you send them the sharing link, and it can be embedded on your website.
Domains
Specify the website domains where you want to embed your chatbot in the text box here.
Usage Limit Settings
Hello Client offers rate limiting to prevent any abuse from users by limiting the number of messages sent from one device on the iframe and chat bubble. By default, it’s set to allow 5 messages every 60 seconds, but you can adjust these values.
Leads
You can prompt a user to input their name, email address and/or phone number when opening the chatbot. The customer information is available to view in the dashboard under the “Leads” tab. You can download them in csv format, along with the conversation history in JSON format.
Chat Interface
You can adjust the user interface of the chatbot and improve the user experience by adjusting the “initial messages”, “suggested messages”, theme, logo, alignment, and colours.
Specifically, to provide clear guidance, you might consider customizing the ‘initial message’ and ‘suggested messages’. The ‘initial message’ serves as your chatbot’s first point of contact with users. This message sets the tone and invites users to engage. The ‘suggested messages’, function as prompts to nudge users towards inquiries. These can be tailored to commonly asked questions or important aspects of your service, such as ‘What is example.com?’ or ‘How does it work?’. In essence, these customizations guide the users’ journey.
- Lead Magnet:
Set up the period of time after which the chat bubble will pop up automatically to engage conversation and capture leads. It is under Settings -> Chat Interface.
- Human Takeover:
This email is specified in the case that visitors want to talk to humans instead of chatbot. By default, it is set up as your login email. It is under Settings -> Chat Interface.
Sources
There are 5 types of data formats for training the chatbot: Files, Text, General Website, Q&A. Every single data format can be used for training. You can also use all different data sources to train your chatbot, but please avoid duplicating content. You need to Retrain Chatbot every time you upload a new data source so that the changes can be updated.
- Files:
All supported files ( .pdf, .doc, .docx, .txt) can be uploaded to knowledge base and trained the chatbot. You need to hit the Retrain Chatbot button after uploading files in order to update your chatbot with new data.
- Text:
The text section is used when you want to amend new data to your chatbot in general. The length of your content can’t be less than 100 characters.
- Website:
Input the url of your target website to scrape the content of website and retrain your chatbot.
- Q&A:
After you add a pair of question and answer to your chatbot, you need to Retrain Chatbot so that the changes can be updated.
Refine the Chatbot’s Instructions
The instructions shapes your chatbot’s behavior and responses. Currently, default base prompt is used to make appointments. To ensure your bot only answers questions about the given document, specify this in the instructions (Settings->Model->Base Prompt).
Utilize the “Revise” Feature and Q&A Data Type
The “revise” feature is accessible from the dashboard in your conversation history. It is a tool for tweaking responses.
If you’re not satisfied with how your chatbot answered a particular query, you can use this feature to alter the response to fix it for the future. Additionally, using the Q&A data type can help your chatbot generate better answers by referring to pre-set questions and answers. The responses you revise will appear in the Q&A tab under “Sources.”
Custom Domain
Use your own custom domains for the embed script, iframe, and chatbot link instead of Helloclient.us
Make Appointment (Calendly)
You need to create a calendly account and authorize Helloclient to access your calendly events:
- Sign Up/Log In: If you don’t have a Calendly account, go to the Calendly website and sign up for a new account.
- Authorization: If you already have an account, authorize Hello Client AI to log in using your credentials.
- Select Event Type
- Write suggested message on chatbot interface to intrigger your calendar
Toll Free Text SMS Integration:
- Apply for a gmail account and a toll free number.
- Authorize Hello Client AI to deploy chatbot to your toll free number.
- Step by Step instruction will be provided to premium subscribers after logging in.
Notification Settings
Your notification settings allow you to receive email updates with chat records from your bot. You can enable notifications by typing an email and clicking “Add Email.”
Next steps
By understanding each setting and adjusting them according to your business needs, you can fully leverage the potential of your Hello Client chatbot. Ensuring your bot is optimized and well-configured will provide your users with a smoother, more intuitive, and effective interaction experience.
Check out our next article for more tips on optimizing user experience with Hello Client.
