Phase 3

• Duration: 6 Sessions
• Sessions: 1 per week (60 minutes each)
• Learning Style: Hands-on, research-driven, project-based
• Capstone Project: Research, build, and deploy an AI assistant solving a real-world problem
• Best For: Students, entrepreneurs, and future AI creators

This course  flows like an AI development journey, ensuring that each student understands every step of the process before building their final project.

Think of this as building AI step-by-step:

• Week 1: Understand AI history and research AI innovations
• Week 2: Learn how AI uses data and structure training data
• Week 3: Train AI to think, speak, and respond effectively
• Week 4: Build and design the AI assistant’s interface
• Week 5: Test, refine, and improve AI responses
• Week 6: Showcase and deploy AI assistant

Goal:

Understand the evolution of AI assistants and research cutting-edge AI applications.

Tools Used:

• Perplexity and DeepSeek – AI-powered research and real-time fact-checking
• OpenAI (ChatGPT) – Conversational exploration of AI history

Key Topics:

• The AI Assistants That Started It All – From Ancient Cultures to ELIZA to Siri and ChatGPT
• AI’s Funniest Mistakes – When chatbots failed spectacularly
• Why AI Assistants Work the Way They Do – The science behind NLP and voice recognition
• Future of AI Assistants – How AI is shaping industries in 2025

Hands-On Activity:

• Use Perplexity to uncover rare AI history facts and create a timeline in ChatGPT.
• Debate Session: "What’s the best AI assistant of all time?"
• Team Challenge: Predict what AI assistants will look like in 2030.

Outcome:

Students will grasp AI evolution and understand how technology impacts AI interactions today.

Goal: Learn how AI assistants use structured and unstructured data and how to prepare AI training sets.

Tools Used:

Perplexity and DeepSeek – Researching how AI models use data

Key Topics:

What is Data? – The difference between structured data (tables, labeled text) and unstructured data (chat logs, audio, images).

Where Does AI Get Its Knowledge? – How AI assistants pull from datasets, APIs, and real-time interactions to generate responses.

Data Cleaning and Labeling – The importance of removing noise, categorizing responses, and structuring datasets for improved accuracy.

Training AI with High-Quality Data – Why "garbage in, garbage out" is a key principle in machine learning and chatbot design.

The Ethics of AI Data – Ensuring privacy, reducing bias, and maintaining transparency in AI training.

Hands-On Activity:

Use Perplexity to analyze how chatbots like ChatGPT source and process data.

Google Sheets Data Exercise:

Sort and label chatbot training data into categories (questions, intent recognition, conversational pathways).

Create "if-then" logic pathways based on structured responses.

Outcome: Students will understand how structured, well-organized data improves AI assistant responses.

Goal: Teach AI assistants how to understand questions, provide meaningful answers, and sound more human-like.

Tools Used:

MindStudio – AI training and dataset management

Re:Tune – Voice and personality training for AI assistants

OpenAI API – Adjusting AI models for better response accuracy

Key Topics:

How AI Understands Language – Breaking down NLP and machine learning

Training AI to Answer Correctly – Fine-tuning prompts for better responses

AI Personality and Voice Training – Making AI sound unique and engaging

Common AI Mistakes and Fixing Them – How AI assistants misunderstand context

Hands-On Activity:

Train an AI model in Open AI and MindStudio using real-world chatbot conversations.

Use Re:Tune to make the AI assistant’s voice more expressive and human-like.

Challenge: Program an AI assistant that can handle at least 20 different user requests.

Outcome: Students will train an AI assistant that can process user queries, understand intent, and respond intelligently.

Goal: Develop a fully functional AI assistant with custom UI and chatbot features.

Tools Used:

Re:Tune – No-code AI chatbot building and UI prototyping

OpenAI API – Integrating AI logic into assistants

MindStudio – Exploring real-world AI chatbot automation

Key Topics:

Designing AI Assistants for Businesses, Schools and Communities

Using OpenAI API to Power Chatbots and Voice Assistants

Customizing UI and Chat Flows with MindStudio

Testing AI Assistants with Real User Inputs

Hands-On Activity:

Build a chatbot interface in MindStudio and connect it to OpenAI API.

Use Intercom’s chatbot testing tools to simulate real user interactions.

Challenge: Develop an AI assistant that can handle customer service inquiries.

Outcome: Each student will have a fully functional AI assistant ready for real-world use.

Goal: Improve AI assistants by testing user interactions and refining responses.

Tools Used:

Open AI – Real-world chatbot testing and feedback loops

MindStudio – Fine-tuning chatbot responses

Re:Tune – Adjusting AI assistant voice behavior

Hands-On Activity:

Run real-world AI assistant tests using Intercom’s simulation tools.

Improve AI assistant’s responses based on real-time user feedback.

Challenge: Refine your AI assistant so it answers 90 percent of user requests correctly.

Outcome: Students optimize AI responses based on real user interaction.

Goal: Deploy AI assistants and present final projects to industry experts.

Final Tools Used:

MindStudio – Showcasing AI chatbot interactions

OpenAI API – Live chatbot demonstration

Re:Tune – AI voice assistant showcase

Outcome: Students launch their AI assistants and present their projects to mentors and industry leaders.

• Duration: Six Sessions
• Sessions: One per week (60 minutes each)
• Learning Style: Hands-on, research-driven, project-based
• Capstone Project: Hands-on, research-driven, project-based.
• Best For: tudents, entrepreneurs, and future AI creators

This course follows a structured AI agent development process, ensuring students master every step before deploying their final project.

Think of this as teaching AI to act on its own:

• Week 1: Introduction to AI agents and automation
• Week 2: How AI agents process and execute tasks
• Week 3: Training AI agents to make smart decisions
• Week 4: Designing workflows and API automation
• Week 5: Testing, refining, and scaling AI agents
• Week 6: Capstone showcase and deployment

Goal: Learn how AI agents operate, execute tasks, and interact with users.

Tools Used:

• OpenAI
• Perplexity 
• DeepSeek 

Key Topics:

• What Are AI Agents? – The difference between AI assistants, agents, and automation.
• The Role of AI in Workflow Automation – How AI speeds up and optimizes business tasks.
• Types of AI Agents – From chatbots to self-learning AI systems.
• AI in the Real World – How companies use AI agents today.

Hands-On Activity:

• Use Perplexity to research the top AI agents being used in business today.
• Breakdown Challenge: Compare human decision-making vs. AI decision-making in common tasks.
• Team Brainstorm: How could AI agents automate tasks in your daily life?

Outcome: Students will understand the basics of AI agents and how they function in real-world applications.

Goal: Learn how AI agents take action based on workflows and automation rules.

Tools Used:

• Make.com
• n8n 
• OpenAI API 

Key Topics:

• How AI Agents Make Decisions – Understanding logic trees and automation workflows.
• Triggering AI Actions – When and how AI executes tasks autonomously.
• APIs and AI Communication – How AI agents connect to external platforms.
• The Balance Between AI and Human Control – Keeping AI collaborative instead of completely independent.

Hands-On Activity:

• Use Make.com to create a simple AI-driven automation flow.
• Test an AI decision model in OpenAI API.
• Challenge: Design a flowchart for an AI agent that handles customer service or scheduling.

Outcome: Students will understand how AI agents process tasks, execute workflows, and automate decision-making.

Goal: Teach AI agents to analyze situations, predict outcomes, and optimize workflows.

Tool Used: Relevance AI 

Key Topics:

• How AI Learns From Data – Reinforcement learning and adaptive workflows.
• Creating AI Decision Pipelines – How AI chains tasks and decisions together.
• Optimizing AI Efficiency – Reducing errors and improving workflow execution.
• Error Handling in AI Agents – Teaching AI to recognize and fix mistakes.

Hands-On Activity:

• Train an AI agent in Crew AI to process multiple inputs.
• Test Relevance AI to deploy a workflow automation model.
• Challenge: Build an AI agent that automates a simple business or personal task.

Outcome: Students will train an AI agent to make decisions and execute multi-step processes independently.

Goal: Create custom automation workflows for AI agents to operate in real-world applications.

Tools Used:

• Make.com
• n8n
• OpenAI API

Key Topics:

• Building AI Workflow Chains – How AI connects tasks in a logical sequence.
• No-Code AI Automation – Creating AI-driven workflows without programming.
• Integrating AI with External Services – Using APIs for business applications.
• AI and Human Collaboration – Designing AI that enhances productivity without full replacement.

Hands-On Activity: 

• Use Make.com to build a real AI automation workflow.
• Test an AI assistant using n8n’s integration tools.
• Challenge: Deploy an AI agent that can automate a business task, such as handling email responses.

Outcome: Students will design AI workflows and automation models using no-code tools.

Goal: Optimize AI agents for efficiency, accuracy, and real-world applications.

Tools Used:

• OpenAI API
• Relevance AI

Key Topics:

• Testing AI Agents in Real Scenarios – Debugging and performance assessment.
• Refining AI Decision Accuracy – Improving AI logic and responses.
• Scaling AI Agents for Larger Workflows – Expanding automation models.
• AI Safety and Ethical Considerations – Preventing errors and unintended consequences.

Hands-On Activity: 

• Run real-world AI agent tests using OpenAI API.
• Improve an AI agent’s performance using Relevance AI.
• Challenge: Optimize an AI agent to complete tasks 50% faster.

Outcome: Students will fine-tune AI agents for better accuracy, performance, and scalability.

Goal: Deploy AI agents and present final projects to industry experts.

Final Tools Used:

• Make.com
• n8n
• Crew AI
• Relevence AI

Outcome: Students launch their AI agents and present their projects to mentors and industry leaders.