0%
Preparing your experience
English 日本語
0 Start Learning

Your Bag

E-learning materials

Building AI Chatbots with LLMs

Design, build, and launch a chatbot on top of a large language model, from system prompt and knowledge base to testing, escalation, and monitoring.

Material Overview
Building AI Chatbots with LLMs

Build a working assistant from start to finish. You will define scope and tone, write a system prompt, connect a knowledge base so answers stay grounded in your own content, and add conversation memory and tool calls for actions such as lookups and bookings. This E-Learning Material also covers handling questions the bot cannot answer, escalation to a human, testing against real conversation logs, and monitoring quality after launch.

Curriculum

Learning Path

Each level is enrolled separately. Work through them in order, or start at the level that matches your experience.

  1. Level 01

    Beginner

    Purpose

    Plan a chatbot before building it by deciding what it will answer and what it will refuse.

    Outcome

    You can specify a chatbot clearly enough that building it becomes straightforward.

    What You'll Learn
    • Define scope, audience, tone, and the questions that belong to a human
    • Write a first system prompt and test it in conversation
    • Collect real questions to use as your test set later

    159 CREDS

  2. Level 02

    Intermediate

    Purpose

    Ground the bot in your own content so answers come from your documents instead of model memory.

    Outcome

    You can build a bot that answers from your material and admits when it does not know.

    What You'll Learn
    • Prepare a knowledge base and split documents so retrieval works well
    • Connect retrieval to the conversation and require citations in replies
    • Handle questions the knowledge base cannot answer without inventing a reply

    387 CREDS

  3. Level 03

    Advanced

    Purpose

    Give the assistant memory and actions so it can complete tasks rather than only answer questions.

    Outcome

    You can deploy an assistant that completes real tasks and hands over cleanly when needed.

    What You'll Learn
    • Add conversation memory and decide what should persist between sessions
    • Connect tools for lookups, bookings, and account actions with permission checks
    • Design escalation to a human with the full context attached

    659 CREDS

  4. Level 04

    Expert

    Purpose

    Operate a production assistant with evaluation, monitoring, and continuous improvement from real conversations.

    Outcome

    You can keep a live assistant accurate and improving after launch rather than leaving it to drift.

    What You'll Learn
    • Score conversations against a rubric and track resolution and escalation rates
    • Detect and fix regressions after a model or prompt update
    • Feed real transcripts back into your test set and knowledge base

    1,035 CREDS