How to Build a Responsible AI Chatbot Using Open Technologies

Lifetime

Intermediate

35 lessons

5 quizzes

0 students

Overview
Curriculum
Instructor

This course provides a step-by-step guide to creating an AI chatbot that provides trustworthy, source-grounded information using curated literature, datasets, and conversational guardrails. All technologies involved are open, lightweight, and accessible.

The knowledge consolidated here is based on our real-life experience of collaboratively developing the Climate Academy Chatbot (CABot). This prototype uses reliable and verified sources to make climate knowledge more accessible to all, while ensuring that the information remains clear, factual and credible.

It was produced in 2026 as part of the AI Learning Labs project through the collaboration between the Climate Academy, #semanticClimate and the Open Knowledge Foundation (OKFN).

Curriculum

7 Sections
35 Lessons
Lifetime

Expand all sectionsCollapse all sections

Introduction
2 Items
- 1.1
  The CABot Story
- 1.2
  How to Do This Course
Module 1 - Understanding AI
9 Items
Large Language Models (LLMs) generate language, not truth. Explore tokens, prompts, context windows and hallucinations, then examine why Retrieval-Augmented Generation (RAG) grounds responses in trusted knowledge.
Module 2 - Building the Knowledge Base
9 Items
Understand how to transform a digital resource (e.g. a book) into semantic knowledge through chunking, embeddings and vector search. Learn why CABot uses ChromaDB and metadata rather than retraining models.
Module 3 - Choosing Your Architecture
5 Items
Compare hosted APIs, local models, vector databases and orchestration frameworks. Rather than recommending one stack, this module helps you choose technologies appropriate to your own community.
Module 4 - Building the Chatbot
8 Items
In this module, we follow a learner's question from a browser prompt to asynchronous RAG orchestration. We aim to understand asynchronous architecture and browser-first deployment.
Module 5 - Responsible & Multilingual AI
5 Items
Responsible AI begins with trustworthy knowledge, not simply powerful models. Explore transparency, source attribution, multilingual retrieval, privacy, accessibility and environmentally conscious AI.
Course Conclusion & Next Steps
2 Items
- 7.1
  Lessons learned
- 7.2
  Recap: Practical workflow for building a responsible chatbot

#semanticClimate

#semanticClimate focuses on semantifying climate data and resources to make climate knowledge open, machine-readable, and actionable. Our Idea is to help scientists, policymakers, and communities respond more effectively to climate change, but more importantly, our aim is to create Youth Knowledge Champions, so that together, we can take Climate Action to each and every citizen in the world. This course content was developed by Renu Kumari on behalf of the entire team who worked on the CABot project (in alphabetical order): Aleena Harold Peter, Ali Mehdi Naqvi, Bui Doan Khanh Linh, Gitanjali Yadav, Parth Kadam, Peter Murray-Rust, Saarthak Patwal, Samyukta Nair, Shreya Bhat, and Udita Agarwal.

Requirements

Foundational Knowledge: Basic familiarity with Python programming, API concepts, and HTML document structures.
Hardware & System: A desktop or laptop computer running Linux, Windows, or macOS.
Software Tools: A standard code editor (e.g., VS Code), Python 3.9+, and a modern web browser.
Local AI Dependencies (Optional for local labs): Ollama installed for running lightweight models locally (e.g., Llama 3.1).

Features

Understand different approaches to building chatbots using Retrieval-Augmented Generation (RAG).
Replicate a chatbot workflow using open, lightweight technologies.
Make informed design decisions around trust, multi-linguality and sustainability.
Build responsible chatbots grounded in curated knowledge.

Target audiences

Data, software, and civic-tech practitioners building responsible AI tools
Climate, environmental, and sustainability educators, researchers, and students
Open knowledge, NGO, library, and community capacity-building practitioners