1. Core Idea
You are not building models—you are choosing the correct model type based on what the solution needs to do.
👉 Ask yourself:
“What capability is required?”
🔑 2. Common AI Model Types and When to Use Them
📝 Text-based Models (NLP)
Use when working with language or text
Capabilities:
Sentiment analysis
Language translation
Text summarisation
Named entity recognition
Example scenario:
Analyse customer reviews to determine if they are positive or negative
✅ Use: Text classification model
🖼️ Computer Vision Models
Use when working with images or video
Capabilities:
Image classification
Object detection
Facial recognition
OCR (text extraction from images)
Example scenario:
Detect defects in manufactured products from images
✅ Use: Object detection model
🗣️ Speech Models
Use when working with audio
Capabilities:
Speech-to-text
Text-to-speech
Speaker recognition
Example scenario:
Convert call centre recordings into text
✅ Use: Speech-to-text model
🔮 Predictive Models (Machine Learning)
Use when making predictions from structured data
Capabilities:
Classification (yes/no, categories)
Regression (numeric prediction)
Example scenario:
Predict whether a customer will cancel a subscription
✅ Use: Classification model
🤖 Generative AI Models
Use when creating new content
Capabilities:
Generate text, code, images
Answer questions (chatbots)
Summarise or rewrite content
Example scenario:
Build a chatbot to answer questions about company policies
✅ Use: Large Language Model (LLM)
⚖️ 3. How to Choose the Right Model (Exam Logic)
Use this quick decision pattern:
Requirement Model Type
Understand or analyse text NLP model
Work with images/video Computer Vision
Process speech/audio Speech model
Predict outcomes from data ML (classification/regression)
Generate new content Generative AI
🧩 4. Key Exam Tip
AI-901 questions are scenario-based, e.g.:
“A company wants to extract text from scanned invoices”
👉 You should recognise:
Input = image
Output = text
✅ Answer: OCR (Computer Vision)
🚀 5. Simple Memory Trick
Think in terms of input type → required capability → model
Text → Understand → NLP
Image → See → Vision
Audio → Hear → Speech
Data → Predict → ML
Prompt → Create → Generative AI
✅ Summary
To identify the appropriate AI model:
Understand the problem
Identify the type of data (text, image, audio, structured)
Match it to the model capability
Select the correct model category