The self-building Chatbot

Core Principles

The FireChatbot Dashboard is designed to give you a comprehensive overview of your Chatbot's performance and allow maintainance of your Chatbot. You can use it to monitor your Chatbot's performance, adapt its design, trigger new Deployments and to manage its Knowledge Base.

The Analytics capabilities of the Dashboard devide into Quantitative and Qualitative Analytics.

Quantitative Analytics

Each time a user enters a page that loads your Chatbot, a page view is sent to the FireChatbot Server. This page view is then used to calculate the number of users that interacted with your Chatbot and the number of messages that were sent to your Chatbot. You can view these numbers in the "Analytics" section of your Dashboard. The Analytics section also provides you with a graph that visualizes the number of users and messages over time. You can use this graph to identify your most visited pages and user engagement over time.

Qualitative Analytics

Every message a user sends to your Chatbot is being analyzed by AI on the FireChatbot Server. The Qualitative Analytics AI is designed to identify the user's intent, categorize the user's message and perform sentiment analysis. There are two sections in the Dashboard, that provide you with Qualitative Analytics data.

Product Interest

The Product Interest Chart shows an aggregated view of the Sentiment Analysis of all user messages. It shows the number of positive, neutral and negative chats, as well as a Fallout Graph, that describes the percentage of Users with different levels of interest towards your products or services. You can use this chart to identify how interesting your website's content is to your users and identify potential issues with the presentation of your content or your products in general.

Questions? Answers!

How do I implement my FireChatbot with a Content Security Policy?

If you use a Content Security Policy on your website, you have to add the following CSP to your website's HTTP header or tag to allow the Chatbot to work properly:

HTTP


                      Content-Security-Policy:

                          default-src 'self'; script-src 'self' https://firechatbot.com;

                          connect-src 'self' https://firechatbot.com;

                          img-src 'self' https://firechatbot.com;

                          style-src 'self' 'unsafe-inline';

                          font-src 'self' https://firechatbot.com;

Do I need to adapt my Terms of Use after I implemented my FireChatbot?

No, you don't have to adapt your Terms of Use after you have implemented your chatbot. FireChatbot comes with integrated Terms of Use, which are fully GDPR compliant and guarantee compliant use.

Which data is stored where?

Knowledge Base

Stores the knowledge, that your Chatbot uses to answer questions

Activated every time a user asks a question

Stores website texts & custom text documents

Provides infinite storage for every Chatbot

Datapoints

[{ id: "Generic ID", type: "website | document", text: "Your website's or document's text, compressed", title: "Your website's url or your document's title", lastUpdate: "Last update of your document or website" },...]

Analytics Database

Stores quantitative & qualitative Analytics data, that is displayed inside the Dashboard

Activated every time a User visits your website

Stores URL, referer, location of the user and counts visits & chatbot activations

Stores chats along with sentiment analysis

Does not generate any IDs or user profiles

Datapoints

Visit Database: [{ chatbotID: "Your Chatbot ID", visits_[data]_[time]: "Number of visits", visits_[data]_[time]_[country]: "Number of visits per country", visits_[data]_[time]_[url]: "Number of visits per URL" },...] Chat Database: [{ chatbotID: "Your Chatbot ID", country: "Country of the user", latitude: "Latitude of the user", longitude: "Longitude of the user", date: "Start date of the chat", endtime: "Time of the last message", messages: ["User and bot messages, excluding images"], page: "URL of the page", probability: "Probability of buying the product (AI generated)", sentiment: "Sentiment of the chat (AI generated)", summary: "Summary of the chat (AI generated)", title: "Title of the chat (AI generated)" },...]

Which AI models is FireChatbot using?

FireChatbot automatically routes a user's question to the AI model, that is best suited to answer it. This process is dependent on the modality of the question (text, audio or image), as well as the complexity. FireChatbot uses GPT-4o, GPT-4o mini, OpenAI Realtime API & Open Source models like Nvidia's Nemotron.

All models are hosted on our own servers in Germany, either via Azure OpenAI Services or directly via Azure.

FireChatbot

FireChatbot | Resources | Get Started

Dashboard & Analytics

Overview

Core Principles

Quantitative Analytics

Qualitative Analytics

Product Interest

Chat Base

Maintenance

Deploy Chatbot

Maintain Knowledge Base

Questions? Answers!

How do I implement my FireChatbot with a Content Security Policy?

Do I need to adapt my Terms of Use after I implemented my FireChatbot?

Which data is stored where?

Which AI models is FireChatbot using?

FireChatbot