Fundamentals of Artificial Intelligence (AI): From Theory to Practice

The programme

Programme overview

Artificial Intelligence is reshaping how organisations operate and compete. This Fundamentals of Artificial Intelligence (AI): From Theory to Practice practical course builds a clear foundation in AI—starting with core concepts and moving quickly into hands-on work, case studies, and simple prototypes. By the end of the week, you’ll understand how the main AI techniques work, where they add value, and how to start implementing them responsibly in your context.

Who should attend

Who is this programme for?

• Business leaders and managers (HR, Finance, Risk, Operations)
• IT professionals
• Software developers
• Data and BI analysts
• Public-sector officials
• Students and graduates exploring AI careers
• Motivated non-technical professionals seeking a solid, practical introduction to AI

Learning outcomes

What you will learn

By the end of this Fundamentals of Artificial Intelligence (AI): From Theory to Practice course, participants will be able to:

• Explain the core ideas behind AI, machine learning, deep learning, and NLP in plain language.
• Select and apply common ML techniques to classification, regression, and clustering problems.
• Build and evaluate simple models using industry-standard tools (e.g., Python, scikit-learn, Keras).
• Turn raw data into usable features and interpret model performance correctly.
• Use cloud and on-prem options to accelerate AI projects.
• Identify ethical, legal (incl. POPIA), and governance considerations and plan for responsible AI use.

Explore the programme

Detailed programme outline

Day 1 — Foundations & impact

• What AI is (and isn’t): narrow vs general AI; where we are today.
• A brief history: from symbolic AI to modern deep learning.
• Where AI delivers value: sector examples (healthcare, finance, retail, public services, supply chain).
• Core toolbox tour: ML, neural networks, NLP, computer vision, automation.
• Lab: “Hello, AI” — run your first notebook; explore a ready-made model; change inputs and interpret outputs.
• Mini-challenge: Spot high-value, low-risk AI opportunities in your organisation.

Day 2 — Machine learning fundamentals

• Supervised vs unsupervised vs reinforcement learning.
• Key algorithms: linear/logistic regression, decision trees & random forests, k-means clustering, SVMs (conceptual).
• Data quality, leakage, train/test splits, cross-validation; metrics (accuracy, precision/recall, F1, ROC-AUC).
• Feature engineering: encoding, scaling, text and date features, handling imbalance.
• Lab: Build a classification model end-to-end (e.g., churn or credit-risk-style dataset), evaluate properly, avoid common pitfalls.
• Mini-challenge: Improve baseline accuracy without overfitting.

Day 3 — Neural networks & deep learning

• Neural network basics: layers, activation functions, loss, optimisation, overfitting.
• Why deep learning matters and when to (not) use it.
• Computer vision with CNNs: image classification workflow.
• Lab: Train a simple neural network (Keras); extend to a small CNN for image classification.
• Mini-challenge: Boost performance with data augmentation/early stopping.

Day 4 — NLP and the AI toolchain

• NLP essentials: tokenisation, embeddings, sequence models, transformers (BERT/GPT concepts).
• Use-cases: sentiment analysis, topic tagging, chatbots, document search, speech basics.
• Platforms & tooling: TensorFlow, PyTorch, scikit-learn; “AI as a Service” (Azure, Google, AWS) – when cloud makes sense.
• Lab options (choose one):Build a simple sentiment classifier, orCreate a lightweight FAQ chatbot using off-the-shelf embeddings.
• Mini-challenge: Compare classical ML vs transformer-based text embeddings on the same problem.

Day 5 — Ethics, governance, deployment & the road ahead

• Responsible AI: bias, fairness, transparency, privacy, security; aligning to organisational policy and POPIA.
• AI governance in practice: roles, documentation, model monitoring, human-in-the-loop, change management.
• From notebook to production: MLOps basics, reproducibility, versioning, model drift.
• Future trends: foundation models, multimodal AI, edge AI; skills planning for teams.
• Capstone: Team project presentations with feedback and next-steps action plans.
• Wrap-up: Personal learning plan, recommended resources, and implementation checklist.

Assessment & certification

• Short daily quizzes, practical labs, and a team capstone.
• Ratomac Certificate of Completion (CPD-aligned hours available on request).

Before you book

Entry requirements

• Prerequisites: Comfortable with spreadsheets and basic maths. No prior coding required (we include a gentle Python primer).

Learning approach

How you will learn

Our diverse instructional approaches ensure effective learning:

– Lectures & Presentations: Engage with expert-driven, stimulating content.– Course Material: Access well-crafted supporting resources.– Group Work: Collaborate on discussions and case studies for practical insights.– Workshops & Role-Play: Participate in immersive, scenario-based activities.– Practical Application: Focus on applying theoretical knowledge in real situations.– Post-Training Support: Receive extensive support after training for skill implementation.

Discuss your requirements

Dates, duration & arrangements

Contact Ratomac to confirm dates, duration, delivery format, entry requirements and any applicable assessment or certification before booking.