AI SOLUTIONS ARCHITECTURE

Get the lay of the land. This session sets the stage — what to expect, how to make the most of the course, and a no-nonsense breakdown of key AI/ML concepts. Let’s kick things off and connect the dots between theory and real-world impact.

• Instructor introduction
• Course overview
• Best practices & recommendations
• Key terms & definitions

Reflection Exercise: Write a brief essay on your AI/ML knowledge, career goals (e.g., AI Architect in Big Tech), and how you plan to build expertise.

Get the big picture of AI/ML — what it solves, how projects unfold, and who’s making it happen. From real-world applications to Generative AI, this session lays the groundwork for navigating the AI-driven future.

• Problem-solving with AI/ML: Regression, classification, process optimisation
• AI/ML project lifecycle
• Roles & project duration
• AI/ML Architect vs. Gen AI Architect
• Generative AI projects: Intro

Assignment #1: Architecture Diagram
Illustrate key components and roles in an AI/ML project.

Building AI models is one thing — keeping them running smoothly is another. We’ll break down MLOps, showing you how to streamline the entire pipeline, from data prep to deployment and retraining. With real-world case studies and a hands-on demo, you'll see why MLOps is the secret to AI that actually works in the wild.

• MLOps: Why a “launch model” approach isn’t enough
• MLOps pipeline overview 
• GenAIOps & its unique pipeline elements
• Case study: Real project examples (Telecoms, Life Sciences, Automotive)
• Demo:  MLOps pipeline walkthrough

Assignment #2: Mini Pipeline Project
Build a simple MLOps pipeline, covering data cleansing, feature engineering, model training, and a mock deployment strategy.

Turn raw data into gold. Master essential data engineering processes — learn how to clean, transform, and prep data using Pandas, PySpark, and Python Ray. Messy data doesn’t belong in your models.

• Data aggregation, ETL, preparation
• Feature engineering, imputation, cleansing
• Tools & libraries: Pandas, PySpark, Python Ray

Assignment #3: Hands-On Data Prep
Clean and engineer features in a real dataset using Pandas or PySpark.

Get a gist of the basics of supervised learning. Get a grip on regression techniques like Linear Regression, XGBoost, and Decision Trees — making predictions shouldn’t feel like guesswork.

• Supervised vs. unsupervised learning
• Regression techniques

Assignment #4: Model Comparison
Solve a regression problem and compare Linear Regression, XGBoost, and Decision Trees.

Cut through the noise of machine learning with classification and clustering essentials. From Logistic Regression to K-Means and Autoencoders, learn when and why to use each approach. By the end, you’ll have a decision-making mind map to tackle real-world data challenges with confidence.

• Logistic Regression
• Clustering & dimensionality reduction
• K-Means clustering & Autoencoders
• Principal Component Analysis (PCA)
• Decision-making mind map 
• Workshop: Thought experiment: Business problem

Assignment #5: Comparative Analysis
Apply classification (e.g., Logistic Regression) and clustering techniques (e.g., K-Means) to a dataset — analyse when to use each.

Get hands-on with model tuning, deployment, and keeping your AI in top shape. Hyperparameters, model chaining, endpoints, and so much more — learn how to fine-tune, launch, and maintain resilient ML pipelines that don’t just work, but keep advancing.

• Model tuning techniques
• Deployment options
• Building robust model pipelines

Deep Learning is the backbone of AI. In this session, you’ll break down deep networks, tinker with reinforcement learning, and get hands-on building your own model in Keras or PyTorch. Time to make AI work for you.

• Deep Learning essentials
• Building your own DL model using Keras or PyTorch
• Reinforcement Learning basics
• Demo: DL model build

Explore how RNNs, CNNs, and Transformers power everything from natural language models to time series forecasting and computer vision. Whether it's decoding text, predicting trends, or spotting objects in images, this lesson gives you the tools to build smarter AI.

• Natural Language Understanding (NLU)
• Innovations in time series analysis
• Computer vision

Get to know the brains behind AI. We’ll break down transformer models, diffusion models, and LLMs — what they are, how they work, and why they matter, focusing on their practical applications and underlying mechanics.

• Transformer model fundamentals
• Decoder-only models, multi-headed attention, autoregression
• Diffusion & reasoning models
• How LLMs work in practice

In this lesson, you’ll dive into the basics of prompt engineering and AI multi-agent systems, all while getting hands-on experience designing prompts and building your very own generative AI chatbot.

• Fundamentals of prompt engineering
• Designing effective prompts 
• AI multi-agent systems & their interactions
• Demo: Building a Generative AI Chatbot

Assignment #6: Prompt Engineering Exercise
Craft a set of prompts and analyse AI-generated responses for a generative task (e.g., summarisation). Start your Capstone Project.

Explore powerful visualisation tools like Matplotlib and Plotly to turn data into insights. Get hands-on building cool AI applications, like a Chat app powered by GenAI, and craft an interactive dashboard with Streamlit to showcase your AI’s potential in style. 

• Common visualisation techniques
• Tools: Matplotlib/pyplot, Plotly, Streamlit
• Building AI applications
• Demo: Interactive dashboard

Get hands-on with building advanced pipelines for large-scale LLMs. Learn to manage models across data, architecture, and dive into distributed training with a focus on mathematical and statistical model integration. 

• Building advanced pipelines for LLMs
• Large-scale & distributed training
• Managing models across data & architecture
• Exploring mathematical & statistical model integration

Lesson 13 takes you on a wild ride through the latest AI breakthroughs. From reasoning models to hardware acceleration and small language model innovations, you’ll get a front-row seat to the future of AI. Get ready to geek out over what's next.

• Reasoning models & Deep Seek
• Hardware acceleration
• Small language models & innovations beyond multi-agents
• Fun facts & recent advancements

AI shouldn’t be a black box. In this lesson, we’ll break down explainability vs. interpretability, tackle bias-variance trade-offs, and explore tools like SHAP and LIME. Question, analyse, and rethink AI ethics in a workshop.

• Explainability vs. interpretability
• Bias vs. variance trade-offs
• Tools & techniques: SHAP, LIME, fairness assessments
• AI ethics
• Workshop: Thought experiment: Bias vs. variance & ethics 

Get hands-on with the cloud platforms shaping AI. From AWS to WatsonX, explore the tools that power AI development — and deploy a model in the cloud while you're at it.

• Overview of major cloud platforms & services
• AWS, Azure, Google, IBM Watson
• Other notable services
• Demo: Cloud deployment walkthrough

Assignment #7: Cloud Services Comparison Report
Compare at least three cloud services for AI/ML deployment — strengths, weaknesses, and best use cases.

Bridge the gap between business and AI with smart, scalable solutions. Learn how to design resilient architectures, optimise for security and cost, and prep for top AI architect roles — portfolio, interviews, and all.

• Approaching a problem
• Functional vs. non-functional requirements
• Designing well-architected solutions
• Interview & portfolio building tips
• Case study: Walkthrough of a complete AI solution architecture

Capstone Project: Choose one of the topics: AI Chatbot for Customer Support, Automated Content Generation for Marketing, AI-Powered Code Assistant, Generative Art & Design Platform, or Personalised Educational Assistant. Create a presentation, a video demo, and a GitHub link of code.