This course develops the programming and data-handling foundation required to build AI applications. Learners work with Python programming, data structures, functions, object-oriented concepts, exception handling, APIs, SQL, notebooks, version control, data analysis and visualisation. Emphasis is placed on writing reusable and testable code, preparing AI-ready datasets and integrating libraries and services into practical workflows. By the end of the course, learners will be able to write and debug Python applications, manipulate structured and unstructured data, perform exploratory analysis and build the programming components required for Machine Learning and AI systems.

Indicative tools and technologies: Python, SQL, NumPy, pandas, Matplotlib, Jupyter, Google Colab and GitHub

This course builds the mathematical reasoning required to understand how AI models learn, optimise and make predictions. Learners explore vectors, matrices, linear transformations, probability, distributions, statistical inference, hypothesis testing, calculus and optimisation. The course connects these concepts with applications such as dimensionality reduction, regression, classification, uncertainty estimation, experimentation and model optimisation. Hands-on exercises help learners interpret model behaviour rather than treating algorithms as black boxes.

Indicative tools and technologies: Python, NumPy, SciPy, SymPy, pandas, Matplotlib and scikit-learn

This course provides an end-to-end understanding of the Machine Learning lifecycle—from problem definition and data preparation to model selection, evaluation and interpretation. Learners build supervised and unsupervised models for regression, classification, clustering, anomaly detection and prediction. The course covers feature engineering, validation strategies, performance metrics, hyperparameter tuning, ensemble methods, explainability and responsible model selection using industry-relevant datasets and use cases.

Indicative tools and technologies: Python, pandas, scikit-learn, XGBoost, LightGBM, imbalanced-learn, SHAP, MLflow and Weights & Biases

This course develops the ability to design and train neural networks for complex AI problems. Learners study perceptrons, backpropagation, optimisation, regularisation, convolutional neural networks, sequence models, attention mechanisms and transformer architectures. Practical work includes building neural networks, improving generalisation, tuning architectures and applying Deep Learning to image, text and sequential data. Learners also examine computational requirements, model efficiency and responsible use of large neural models.

Indicative tools and technologies: PyTorch, TensorFlow-Keras, Torchvision, OpenCV, TensorBoard, Optuna, MLflow and Weights & Biases

This course enables learners to build applications that understand, retrieve and generate human language. It begins with text preprocessing, classification, sentiment analysis, topic modelling and embeddings before progressing to transformers and Large Language Models. Learners work with prompt engineering, semantic search, vector databases, Retrieval-Augmented Generation, LLM evaluation, guardrails and responsible use of generative models. The focus is on developing reliable LLM applications grounded in enterprise documents and domain knowledge.

Indicative tools and technologies: Python, spaCy, NLTK, Hugging Face Transformers, Sentence Transformers, LangChain or LlamaIndex, FAISS or Chroma, FastAPI, Streamlit and Gradio

This course develops skills in extracting intelligence from images and video and generating new visual content. Learners study image processing, convolutional networks, image classification, object detection, segmentation, feature extraction and video analytics. The Generative AI component introduces autoencoders, Generative Adversarial Networks, diffusion models, image generation and multimodal visual applications. Learners build practical solutions while considering bias, privacy, authenticity and responsible use of synthetic media.

Indicative tools and technologies: Python, OpenCV, PyTorch or TensorFlow, Torchvision, YOLO frameworks, Hugging Face Diffusers, Matplotlib and MLflow

This course focuses on building AI systems that can plan, use tools, retrieve information, maintain context and execute multi-step tasks. Learners explore agent architectures, tool calling, memory, reasoning workflows, orchestration, multi-agent collaboration and human-in-the-loop controls. Applications include research agents, enterprise knowledge assistants, workflow automation, customer-support agents and decision-support systems. The course also addresses agent evaluation, observability, security, access controls and responsible deployment.

Indicative tools and technologies: Python, APIs, LangChain, LangGraph, Microsoft AutoGen or equivalent frameworks, vector databases, FastAPI and workflow automation platforms

Capstone Project I enables learners to apply their first-year learning to a real-world business, industrial or social problem. Learners identify a problem, conduct a literature review, define the proposed solution, prepare the data, select suitable models and develop a working prototype. Each project is supported through structured mentoring, periodic reviews, documentation and a formal presentation. The capstone demonstrates problem-solving ability, technical implementation, communication and the capacity to convert an AI idea into a credible proof of concept.

Typical outputs: Problem statement, literature review, dataset and methodology, baseline model, working prototype, technical report, presentation and viva voce.

This course introduces AI systems that learn through interaction and improve decisions based on rewards and feedback. Learners study agents, environments, states, actions, policies, Markov Decision Processes, Q-learning, policy-based methods and Deep Reinforcement Learning. The systems-design component examines how AI solutions should be architected for scalability, reliability, latency, security, explainability and cost. Learners connect algorithms with the engineering choices required to build dependable intelligent systems.

Indicative tools and technologies: Python, Gymnasium, Stable-Baselines3, RLlib, PyTorch or TensorFlow, simulation environments and system-design tools

This course prepares learners to build AI applications that combine speech, text, images and documents. Learners work with speech recognition, text-to-speech, language understanding, conversational design, chatbots, voice assistants and multimodal retrieval. Practical work includes integrating speech and language models, creating multimodal knowledge assistants and evaluating human-AI interactions for relevance, safety and user experience.

Indicative tools and technologies: Python, Whisper or equivalent speech models, Librosa, Torchaudio, Hugging Face Transformers, Rasa or Botpress, LangChain or LlamaIndex, vector databases, Streamlit and Gradio

This Microsoft Azure certification-aligned course develops the ability to design, build, manage and deploy AI applications and agents using Azure AI services and Microsoft Foundry. Learners work with generative AI applications, agentic solutions, Azure AI Search, document intelligence, language, speech, vision, Retrieval-Augmented Generation, content safety, security, evaluation and responsible AI. Practical exercises focus on integrating Azure services into enterprise-ready applications using Python, APIs and SDKs.

Indicative tools and technologies: Microsoft Foundry, Azure AI Services, Azure AI Search, Azure OpenAI patterns, Azure Document Intelligence, Azure SDK for Python and Azure application services

This Microsoft Azure certification-aligned course develops the skills required to operationalise Machine Learning and Generative AI systems. Learners study MLOps infrastructure, automated training pipelines, model deployment, CI/CD, monitoring, governance, versioning and lifecycle management. The course also introduces GenAIOps practices for deploying, evaluating, tracing, monitoring and optimising generative AI applications and agents. Learners build secure and repeatable workflows that move models from experimentation into reliable production environments.

Indicative tools and technologies: Azure Machine Learning, Microsoft Foundry, GitHub Actions, MLflow, Azure CLI, model registries, monitoring, observability and infrastructure-as-code concepts

Capstone Project II requires learners to design and implement an advanced end-to-end AI solution with greater technical depth and deployment readiness than the first capstone. Learners conduct experimentation, compare models, validate results, test system performance and address scalability, reliability, security and responsible AI considerations. The final solution should demonstrate clear technical or business value and be presented through a working demonstration, formal report and viva voce before an expert panel.

Typical outputs: Production-oriented AI solution, experimental results, evaluation framework, deployment architecture, project repository, final report, demonstration and viva voce.

This course helps learners convert their applied AI work into a structured research contribution. Learners refine the research problem, conduct a systematic literature review, articulate the methodology, analyse experimental results and document the originality and limitations of their work. The course includes research writing, referencing, similarity checks, research ethics, mentor review, journal or conference selection, submission preparation and research presentation. The objective is a publication-ready paper connected to the learner’s capstone work.

Typical outputs: Research manuscript, supporting experimental evidence, similarity report, mentor-reviewed submission and research presentation.