Day 15 – Bonus: Build an AI Chatbot Lesson

Lesson Overview

This bonus lesson shows a practical way to put AI to work on your website. Using Chatbase, you will build a chatbot that answers questions using your own content. The bot reads from uploaded files or linked sources and uses those materials as its knowledge base when visitors ask questions. This is useful for handling customer service, presales questions, and basic onboarding without constant human involvement.

The lesson focuses on the essentials. You will see how to gather and upload files, review model options, and write simple instructions that guide the bot’s behavior. You will also preview design options for the chat interface and learn how to publish the bot and embed it on a page. The free plan is limited, but it is enough to understand the workflow. A paid plan gives more message credits and a much larger character limit, which is helpful when you need a sizable knowledge base.

This fits the boot camp by connecting earlier model discussions to a real deployment. If you already have a website and a set of FAQs, product docs, or help articles, you can make those materials instantly useful to visitors through an AI chat window.

Who This Is For

If you have a website and want AI to handle repetitive questions, this lesson will help you ship something useful quickly. It supports both first-time builders and those ready to scale.

• Small business owners who need a quick help desk for customers
• Course creators who want a tutor-style chat that answers from lesson materials
• Marketers who need to answer pricing and feature questions on launch pages
• Support teams that want to deflect common tickets with accurate answers
• SaaS founders who want a simple presales and onboarding companion
• Educators who want a reference bot for class resources and policies

Where This Fits in a Workflow

Use this lesson after you have content worth answering from, such as FAQs, product sheets, tutorials, or policy docs. Organize those files, export them as PDFs if they live in Google Drive, and upload them as your knowledge base. Then set the bot’s tone and fallback behavior so it answers from your materials and clearly declines when it lacks information.

Common applications include placing the bot on a pricing or features page to handle presales questions or embedding it on a support page to reduce ticket volume. Once the bot is live, check the activity logs to see real questions and spot content gaps. Update your files or instructions as needed, then republish.

Technical & Workflow Benefits

Manual support and sales chats require constant monitoring and can repeat the same answers all day. With a Chatbase bot, you load your documents once, set clear behavior, and let the bot respond using that content. The default model option is GPT-4, which is a strong general choice, and you can test answers in the playground before embedding.

A practical difference shows up immediately on pages that get frequent questions. For example, a product page that triggers “What does this feature do?” can answer directly by referencing your uploaded PDFs. On a help page, the bot can field setup questions pulled from your how-to documents. The free plan gives you a feel for the flow. A paid plan adds more message credits and a much larger character limit, which supports bigger knowledge bases and more traffic. The result is faster responses, fewer repeated tickets, and consistent answers shaped by your own content.

Practice Exercise

Try a small, realistic build to see the end-to-end workflow.

Scenario: Pick one topic you already document, such as a product, class, or tool. Gather 3 to 4 files that explain the basics, FAQs, and key features. If the files live in Google Drive, export them as PDFs.

Steps:

1. Create a new Chatbase bot and upload your PDFs. Keep the total size modest for this exercise.
2. In AI settings, keep GPT-4 selected. Write simple instructions such as: “Use a casual, friendly tone. Always answer from the knowledge base. If you do not find an answer in the documents, say you do not have that information.”
3. Test in the playground. Ask three questions your visitors often ask. Adjust files or wording until the answers are clear.
4. Publish: Make the bot public, copy the iframe code, and embed it on a draft page in Framer, Wix, Squarespace, or WordPress.

Reflect: Which questions were answered well, and which exposed gaps in your documents or instructions?

Course Context Recap

This bonus lesson adds a practical deployment to the 14-Day AI Boot Camp. Earlier videos discussed model choices. Here you see how those choices show up in a real chatbot that runs on your own materials. The video walks through creating the bot, writing effective instructions, testing responses, and embedding it on a website. Use this to offload basic customer service and sales questions, then build on it by adding more documents and refining your instructions. Continue through the boot camp to round out your AI skill set and apply what you build to real projects.