TECHCRUSH AI/ML
BOOTCAMP
Facilitator: Lamzey Aribido
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
GROUND RULES
HAVE AN ASK
PUNCTUALITY
OPEN MIND QUESTIONS
PARTICIPATION ONE CLASS RESPECT
YOUR MIC MUST RAISE YOUR
BE MUTED HAND BEFORE
ALWAYS YOU SPEAK
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
AI/MACHINE LEARNING TOOLS
Keras Google Colab NumPy
Pandas Python Scikit Learn
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
Practical Real-
World Scenario
WELCOME,
ORIENTATION,
&
RECAP & MINI
INTRODUCTION
TO PROJECT
AGENDA
ARTIFICIAL
INTELLIGENCE
Introduction To Artificial Intelligence
TOOLS SETUP
&
PYTHON RECAP
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
DAY 1: WELCOME, ORIENTATION, & INTRODUCTION TO ARTIFICIAL INTELLIGENCE
What is AI?
At its core, Artificial Intelligence is the capability of a machine or system to
imitate intelligent human behavior. It involves creating systems that can
automate complex reasoning, perceive their environments, and make
decisions. AI is not magic; it is applied mathematics and logic.
Why AI now?
While the concepts of AI have existed since the 1950s, three recent
breakthroughs have caused the current explosion in the field:
• Massive Data Availability: We generate more digital information today
than ever before.
• Advanced Compute Power: Processing hardware has become
exponentially faster and cheaper.
• Algorithmic Efficiency: We can now compress complex models to run
locally on minimal data payloads, bringing AI directly into everyday
devices and edge environments.
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
DAY 1: WELCOME, ORIENTATION, & INTRODUCTION TO ARTIFICIAL INTELLIGENCE
The AI Ecosystem: AI vs. ML vs. DL
Think of these concepts as a set of nested layers:
• Artificial Intelligence (The Outer Layer): Any technique that enables a
computer to mimic human intelligence. This includes simple rule-based
automation (like a thermostat turning on when the temperature
drops).
• Machine Learning (The Middle Layer): A subset of AI where
algorithms are not given hardcoded rules. Instead, they parse historical
data, learn the underlying patterns, and make informed decisions
independently.
• Deep Learning (The Inner Layer): A highly specialized subset of
Machine Learning that uses multi-layered artificial neural networks.
These complex networks are designed to process massive amounts of
unstructured data, like recognizing faces in live video or understanding
spoken language.
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
DAY 1: WELCOME, ORIENTATION, & INTRODUCTION TO ARTIFICIAL INTELLIGENCE
The AI Workflow
Every successful AI project follows a fundamental, three-
step pipeline:
1. Data (The Fuel): Gathering, cleaning, and preprocessing
data. This step is often 80% of the work. The focus here is
on data minimization—collecting high-quality, relevant
signals rather than useless noise.
2. Model (The Engine): This is where your chosen
algorithms are trained to recognize patterns within that
carefully prepared dataset.
3. Deployment (The Vehicle): Taking the trained model
out of the lab and putting it into the real world.
Deployment can mean hosting an API on a cloud server or
embedding a compressed AI model directly onto a tiny
microcontroller to run automated feedback loops locally.
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
DAY 1: WELCOME, ORIENTATION, & INTRODUCTION TO ARTIFICIAL INTELLIGENCE
Real-World Applications
AI is currently transforming every major industry. Here is how it is being applied today:
• Business: Predictive analytics for customer churn, automated support agents, and dynamic pricing models.
• Health: Computer vision for medical imaging—such as detecting anomalies in X-rays faster than the human eye—and personalized
treatment planning.
• Agriculture & Aquaculture: Localized sensor analytics. For example, AI can monitor a catfish pond's telemetry (like pH and oxygen
levels) and automatically trigger feeding systems. In rural environments, these systems often utilize "Offline-First" computing to function
without relying on internet connectivity.
• Smart Infrastructure: Intelligent traffic management frameworks utilizing vehicle-to-everything (V2X) communication. AI analyzes
vehicle flow at intersections to adjust lights dynamically, reducing congestion.
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
DAY 2: TOOLS SETUP & PYTHON RECAP
Setting Up Our Environment: Google Colab.
Before we can build intelligent models, we need a reliable workspace. In this bootcamp, we will primarily
utilize Google Colab.
Why these tools?
• Zero Setup: Google Colab runs entirely in the cloud. You do not need to worry about configuring complex
local environments or lacking computational power on your personal machines.
• Interactive Coding: Both tools allow you to write and execute Python code in modular "cells." This means you
can run a small piece of logic, see the immediate output, and debug on the fly without running an entire
massive script.
• Documentation: Notebooks allow you to seamlessly blend executable code with formatted text (Markdown),
making it easy to take notes directly alongside your functioning models.
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
DAY 2: TOOLS SETUP & PYTHON RECAP
Python Recap: The Language of AI Loops
Python is the industry standard for Artificial Intelligence Loops allow us to automate repetitive tasks, such as
iterating through a stream of data points.
because of its readability and the massive ecosystem of
• For Loops: Used when iterating over a sequence (like a list
data science libraries available. Let's quickly review the
of historical readings).
core building blocks you will use daily.
Variables & Syntax
Variables are containers for storing data values. In
Python, you do not need to declare the type explicitly.
While Loops: Used to execute a set of statements as long as
a condition remains true. This is heavily used in continuous
control loops.
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
DAY 2: TOOLS SETUP & PYTHON RECAP
Installing Common Libraries
We do not need to write complex algorithms from scratch. Python has powerful libraries designed specifically for the AI workflow. You can
install or import these directly into your notebooks.
• NumPy (Numerical Python): The foundational library for mathematical computing. It allows us to process large arrays and matrices of
numerical data incredibly fast.
• Pandas: The go-to tool for data manipulation and analysis. It allows you to import datasets (such as CSV files containing months of logged
data) and organize them into structured tables called DataFrames for easier cleaning.
• Scikit-learn (Sklearn): The premier library for standard Machine Learning. It contains prebuilt algorithms for classification, regression, and
clustering, along with tools to evaluate how accurately your models perform.
How to install in a notebook environment:
(Note: Colab has these pre-installed, but you can always ensure you have the latest versions using the pip package manager)
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
DAY 3: RECAP & MINI PROJECT
Recap Quiz and Discussion
Today, we consolidate everything we have covered so far. Before we build our first automation script, we will review the core concepts
from Day 1 and Day 2.
Key Review Topics:
• The AI Landscape: Differentiating between standard rule-based Artificial Intelligence, Machine Learning, and Deep Learning.
• The Workflow: Understanding the Data à Model à Deployment pipeline.
• Python Basics: Ensuring everyone is comfortable with variables, loops, and our Google Colab environment.
(Instructor Note: Use this time to open the floor for Q&A, addressing any installation or conceptual hurdles before moving into practical
application.)
Disclaimer: This training material belongs to techcrush and shouldn’t be shared
DAY 3: RECAP & MINI PROJECT
Mini Project: Hello AI World
Objective: Build a basic automation script utilizing
decision logic.
Before we train models that "learn" from data, we
must understand how to program machines to make
automated decisions based on incoming data. In this
mini-project, we will write a Python script that
simulates a localized edge sensor continuously
Your Task:
monitoring an environment and triggering an action
1. Copy the starter code into a new cell in your Colab Notebook.
when a specific threshold is met.
2. Run the cell and observe the output.
Scenario: We are simulating a smart environmental 3. Modify the script: Change the scenario. Instead of temperature, adapt
sensor. It will read simulated telemetry data (e.g., the logic to monitor something else (e.g., a customer's bank balance, vehicle
temperature) and use decision logic to trigger an speed, or water pH levels). Add an elif statement to create a "Warning" state
alert if the conditions become unsafe. before the full "Alert" state is triggered.
Disclaimer: This training material belongs to techcrush and shouldn’t be shared