Desktop Image 1400x788 16 9

Advanced Program in AI and Machine Learning with Agentic AI Systems

Work Experience

Starts On

Duration

8 Months Online + Live Online

8-10 hours per week

Program Fees

Applicable Taxes will be charged at checkout

ELIGIBILITY

Bachelor’s degree or higher

Knowledge of mathematics and statistics required

Future-Proof Your Career

Emeritus and the University of Melbourne unite to offer an exclusive opportunity. Enroll by Invalid liquid data and get a 5% by using code: APAC5PERCENT9622 program fee benefit. Limited seats available. Reserve your spot today!

What is the Advanced Program in AI and Machine Learning with Agentic AI Systems about?

With AI moving from experimentation to enterprise-grade applications, advanced AI engineering skills are becoming essential. As organizations embrace AI-driven transformation, demand is growing for professionals who can build, deploy, and scale intelligent, production-ready AI systems responsibly.

The Advanced Program in AI and Machine Learning with Agentic AI Systems by the University of Melbourne equips professionals with in-demand skills across Generative AI, LLMs, RAG, and Agentic AI. Through hands-on implementation, learners explore AI agents, develop intelligent applications, and build scalable solutions using modern AI engineering and MLOps practices.

By combining Deep Learning, Transformers, Reinforcement Learning, Agentic AI, and Responsible AI, the program prepares professionals to tackle real-world challenges across industries. Global faculty expertise, practical assignments, optional IBM certifications, and handson tool-based learning further strengthen technical capabilities, complemented by a qualification from a globally top-ranked university.

88%

AI adoption is widespread
Source: Stanford HAI, AI Index Report 2026

53%

AI agents are automating business processes.
Source: Microsoft, 2025 Work Trend Index

47%

Gen AI is moving into production
Source: EY India, The AIdea of India 2025

Why Choose The University of Melbourne

#1

University in Australia
Source: Average of the latest ARWU, QS, and Times Higher Education Global Rankings

#1

in graduate employability in Australia
Source: The Global University Rankings 2026

#22

in the world
Source: QS World University Rankings 2027

What are the highlights of the Advanced Program in AI and Machine Learning with Agentic AI Systems?

Learning Journey Orientation immersion

Learn at your own pace through live expert-led sessions and on-demand videos.

Learning Journey Learn at your own pace

Gain insights from UoM faculty through masterclasses on emerging AI advancements.

Learning Journey Build real-world skills

Build real-world skills with 15+ industry-relevant tools and 30+ use cases.

Learning Journey Live learning from the best

Explore advanced AI technologies, including RAG, Agentic AI, AI Agents and MLOps.

Learning Journey IBM certifications

Earn three optional IBM certifications in advanced Generative AI applications.

Learning Journey Capstone project

Develop practical AI/ML expertise through 20+ hands-on projects.

Learning Journey Final campus immersion

Build a professional GitHub portfolio showcasing your code and projects.

Learning Journey Career support

Apply your learning through an end-to-end AI capstone project.

Decode global frameworks in machine learning and AI

Decode global frameworks data science

Understand and apply AI/ML fundamentals to real-world problems. 

Decode global frameworks Core AI and ML implementation

Perform data analysis and feature engineering using Python and SQL.

Decode global frameworks Natural language processing

Evaluate and interpret machine learning models across supervised and unsupervised learning techniques.

Decode global frameworks Vision and speech systems

Implement advanced machine learning techniques including ensemble modelling, recommendation systems, and time-series forecasting 

Decode global frameworks Reinforcement learning

Apply deep learning techniques using neural networks, CNNs, RNNs, and Transformers. 

Decode global frameworks Model deployment and tools

Understand and implement Generative AI techniques including GANs, VAEs, and diffusion models. 

What makes this program so unique IBM certifications

Develop NLP and LLM applications using embeddings, prompting, and model adaptation techniques. 

Decode global frameworks Model deployment and tools

Design and implement RAG and Agentic AI systems using retrieval, vector databases, and agent architectures

Decode global frameworks Core AI and ML implementation

Deploy and manage AI solutions using MLOps, monitoring, containerisation, and CI/CD practices

What makes this program so unique faculty masterclasses

Evaluate AI systems for responsible and secure deployment, considering fairness, explainability, privacy, governance, and risk. 

Who is this designed for?

  • Data Scientists, Data Analysts, ML/AI Professionals: Looking to Deepen their knowledge across modern AI, ML, Generative AI, RAG, and Agentic AI.

  • Software Engineers, Developers, Data Engineers, IT & Technology Professionals: Looking to Build practical understanding of modern AI technologies, and their application in technology solutions.

  • Product Managers, Business Analysts, Technology Consultants, Solution/Technical Architects, Technical Project & Program Managers: Aiming to Build AI fluency to work effectively on AI-enabled products, solutions, architecture, and technology initiatives.

 

An industry-leading curriculum for all your skill needs

This rigorous and application-focused curriculum is delivered in multiple pillars, blending foundational concepts with hands-on upskilling in core technical areas of generative AI and ML.
  • What is Artificial Intelligence?

  • Understanding Machine Learning

  • Types of Machine Learning

  • The Broader Skills Ecosystem

  • The Machine Learning Workflow

Learning Outcomes

Differentiate AI, ML, and Deep Learning, frame business requirements as measurable AI problems, map high-level solution architecture, identify relevant personas, and run a minimal end-to-end AI application.

  • Vectors, Scalars, Matrix, Operations on Matrix, Determinants, Role of stats in DS, Types of data, Descriptive Stats, Intro to Probability, Probability Distributions

  • Inferential Statistics, Sampling, Estimation, Hypothesis, Type 1 and Type II errors, Z test, T test, Z score, and Confidence Interval

Learning Outcomes

Apply core mathematical, and statistical concepts to machine learning, including feature representation, uncertainty, hypothesis testing, and model evaluation.

  • W1- Variables and data types · operators · strings · input/output · Lists, Tuples, Sets, Dictionaries · conditionals · loops · functions and lambda

  • W2- Comprehensions · generators · exception handling · OOP (classes, inheritance, polymorphism, encapsulation) · modules and packages · datetime

  • W3- Excel (importing, cleaning, grouping, pivots) · SQL (SELECT, aggregates, GROUP BY, JOINs) · IDEs, Jupyter, Colab · Git (repos, commits, branching, GitHub Desktop, collaboration)

  • W4- NumPy (arrays, vectorisation, 2D arrays) · Pandas (Series, DataFrame) · loading and inspecting data · missing values · duplicates and type conversion · outlier detection (IQR, Z-score)

  • W5- Descriptive statistics and distributions · correlation and cross-tabulation · Matplotlib and Seaborn · univariate and bivariate analysis · heatmaps and pairplots · train-test split and cross-validation intro

Learning Outcomes

Write modular Python programs, analyse, and visualise structured data, use Git for source control, and collaboration, and build reproducible data-analysis workflows.

  • Data cleaning, feature selection, and normalization

  • Hands-on exercises, case studies, and discussions

Learning Outcomes

Profile datasets, apply data cleaning, and preprocessing techniques, engineer, and select features, and construct leakage-safe preprocessing pipelines.

  • Linear regression and evaluation metrics

  • Multiple, Polynomial, Overfitting, Solution to Over Fitting

Learning Outcomes

Profile datasets, apply data cleaning, and preprocessing techniques, engineer, and select features, and construct leakage-safe preprocessing pipelines.

  • Evaluation metrics

  • Logistic regression, Decision tree, Random Forest, SVM, Model Deployment basics (store, load, predict)

Learning Outcomes

Formulate classification problems, train, and compare classification models, evaluate their performance, tune hyperparameters, and serialize models for reproducible inference.

  • Basics, Distance Metrix and Applications

  • How to implement clustering (Agglomerative clustering) and connect with business requirements, Algorithms (PCA, CFA), Association Rule Mining, Algorithm, DB Scan, Anomaly Detection (Nearest Neighbor, and Isolation Forest)

Learning Outcomes

Select appropriate similarity measures, apply clustering, dimensionality reduction, and association-rule mining, and detect anomalies in business use cases.

  • Ensemble technique with examples (its difference from supervised and unsupervised learning); types (Bagging and boosting )

  • Bagging and boosting and different algorithms, Libraries (Adaboost, GB, XGB, Catboost)

Learning Outcomes

Explain, and apply ensemble techniques such as bagging, and boosting, compare ensemble models, and evaluate their impact on generalisation.

  • Similarities - how to measure (PSN correlation, Cosine), Distances (Manhattan, Euclyedian), and Use cases

  • What is recommendation engine, Its Purpose, Types and how to build (libraries)

Learning Outcomes

Compute similarity measures, differentiate recommendation approaches, evaluate recommender systems, and build, and assess a hybrid recommendation pipeline.

  • Types of timeseries data, AR and MA Modelling

  • ARIMA, FB Prophet, and implement data

Learning Outcomes

Analyse time-series patterns, engineer forecasting features, fit, and interpret forecasting models, and evaluate them using time-aware validation, and metrics.

  • Perceptrons, math behind perceptrons, and Python implementation

  • Introduction to MLPs, forward propagation, Python implementation, Introduction, math derivation, and Python implementation

  • Cross-validation, neural networks, activation functions, and DL frameworks

  • Cross-validation, neural network coding, and applications

Learning Outcomes

Select appropriate similarity measures, apply clustering, dimensionality reduction, and association-rule mining, and detect anomalies in business use cases.

  • Introduction to optimizers, activation functions, Loss Functions, Overfitting Scenario

  • Best Practices in choosing optimizers, activation functions, loss function, batch normalization and dropout technique

Learning Outcomes

Interpret optimisation concepts, compare optimisation techniques, apply regularisation, and use training/validation curves to diagnose convergence.

  • Convolution Neural Network (Filters - Feature Detectors, Pooling - Avg, Max, Padding and Stride), Basic Architecture

  • Pre-trained Networks, Transfer Learning and Fine Tuning

  • Recurrent Neural Networks (Temporal Nature of Data, Recurrent Mechanism, Types of RNN), LSTM Gates, and GRU Gates

  • Applications, Drawbacks of RNN, LSTM and its drawbacks, GRU. Attention Mechanism, and Transformers

Learning Outcomes

Understand CNN, and recurrent network concepts, build, and evaluate basic CNN, and LSTM/GRU models, and explain attention as a bridge to Transformers.

  • Autoencoders (Vanilla, Convolutional, Denoising), VAE, Self-Supervised Learning, Anomaly Detection + Applications for various use cases

Learning Outcomes

Understand autoencoders, VAEs, and self-supervised learning approaches; inspect learned representations and implement reconstruction-error-based anomaly detection.

  • VAE, GAN, Architecture, Training Process of Generator and Discriminator, DCGAN, cGAN, CycleGAN, StyleGAN, Diffusion Models, Stable Diffusion (Text-to-Image), ControlNet Implementation, Application, and Evaluation Metrics (FID, IS)

Learning Outcomes

Differentiate generative, and discriminative AI, build a VAE, explain GANs, and diffusion models, and evaluate generative outputs using FID, and IS.

  • Introduction to NLP, Text Preprocessing, Tokenisation (Word, Subword, Character), Text Representation (BoW, TF-IDF), Word Embeddings (Word2Vec, GloVe, FastText)

  • LSTM Text Classification, Static vs Contextual Embeddings, Sentence Encoders, Semantic Search, Bridge to Transformers

Learning Outcomes

Apply NLP preprocessing, and text-representation techniques, build a baseline LSTM-based text-classification workflow, and implement or explain semantic-search workflows.

  • Introduction, Transformer architecture, Attention Mechanism, and Types (GPT, BERT, T5, etc.)

  • Machine Translation and Text Generation

Learning Outcomes

Understand Transformer architecture and attention mechanisms, differentiate major Transformer families, trace Transformer data flow, and apply Transformer models to text generation and machine translation.

  • LLM - definition, Pretraining objectives, fine-tuning, transfer learning, prompt engineering, applications, Other LLMs, and Embedding

  • Prompt Engineering, Agents, and Vector DBs

Learning Outcomes

Explain LLM fundamentals, compare prompting, RAG, and fine-tuning as adaptation strategies, explain their role in LLM applications, and design a basic LLM application flow.

  • RAG Architecture, RAG frameworks, Use of FAISS/ChromaDB, vector search, Core concepts of LLM-powered Agentic AI, Agent Architecture, Hands-on agent design,

Learning Outcomes

Design RAG pipelines, compare, and apply retrieval approaches, evaluate retrieval, and generation, explain agent architecture, and build an end-to-end RAG/agentic prototype.

  • Introduction to RL, Multi-Armed Bandits, Markov Decision Process, Rewards, Returns & Discounting, Policies & Value Functions, Dynamic Programming, Monte Carlo, TD Learning, Q-Learning & Bellman Update, Exploration (Epsilon-Greedy, SARSA), Reward Design, Algorithm Landscape, Evaluating RL Agents / Implementation and Hands-on Labs

Learning Outcomes

Formulate sequential decision problems as MDPs, compute, and interpret key RL concepts, compare foundational RL methods, and implement Q-Learning, and SARSA.

  • Why Deep RL

  • Deep Q-Networks (DQN)

  • DQN Improvements (Double, Dueling, Prioritised Replay)

  • Policy Gradient Methods (REINFORCE)

  • Actor-Critic, PPO & SAC

  • RLHF & Learning from Preferences (GRPO)

  • Algorithm Selection

  • Training Curve Diagnosis

  • Evaluating DRL Agents

  • Implementation and Hands-on Labs

Learning Outcomes

Explain major deep reinforcement-learning approaches, differentiate algorithm families, select appropriate algorithms, and evaluate DRL agents.

  • Why MLOps & Silent Failure

  • ML Lifecycle & Deployment Patterns

  • Experiment Tracking & Model Registry (MLflow)

  • Drift Detection (Evidently)

  • Docker & Containerisation

  • Kubernetes (Pods, Deployments, Scaling)

  • Monitoring (Prometheus & Grafana)

  • CI/CD Pipelines & Quality Gates

  • Implementation and Hands-on Labs

Learning Outcomes

Explain the production ML lifecycle, track experiments, and model versions, package inference services using Docker, implement a basic CI/CD quality gate, interpret drift reports, and support reproducibility, and auditability.

  • Responsible AI principles (Fairness, Transparency, Accountability); Bias and mitigation; Model Explainability (LIME, SHAP); APIs for transparency and auditability; Data privacy (GDPR, DPDP); Security risks and red teaming; AI regulations (EU AI Act, ISO); Ethical concerns in Generative AI (misinformation, consent, copyright)

Learning Outcomes

Explain Responsible AI principles, identify sources of bias, analyse AI security, and abuse risks, map AI systems to governance requirements, and conduct a structured Responsible AI assessment.

  • Final project across concepts

Learning Outcomes

Integrate program concepts to solve a real-world AI problem, and design, develop, evaluate, and deploy an end-to-end AI solution.

  • Basic Image Processing Techniques, Image Scaling, Object Detection Algorithms, and Image Segmentation Algorithms

  • Video Analysis, Self-Supervised Learning techniques, and Few Shot Learning

Learning Outcomes

Understand core computer vision tasks, and the computer vision pipeline; apply preprocessing, transfer learning, and augmentation techniques; analyse object detection, and segmentation approaches; evaluate vision models;, and assess deployment, optimisation, and ethical considerations.

Note: The program curriculum is subject to change to reflect evolving industry requirements, emerging technologies, and market trends.

Github Portfolio

Get an edge in the market with your digital portfolio on GitHub. Share code, and collaborate with other enthusiasts on projects that add credibility to your resume. Additionally, you will also
Create your own profile or optimise your existing one

Create your own profile or optimise your existing one

Publish 2 projects (including the capstone project) on your GitHub portfolio

Publish 2 projects (including the capstone project) on your GitHub portfolio

Get insights from domain experts

Get insights from domain experts on how/why GitHub is a key differentiator in the interview process

Tools covered

IBM certifications

IBM certifications

Enhance your credentials, and practical expertise with three optional IBM certifications focused on Multimodal Generative AI, AI Agents with RAG, and LangChain, and Responsible AI.

  • Certification I: Developing Multimodal Generative AI Applications 

  • Certification II: Building AI Agents with RAG and LangChain 

  • Certification III: Responsible and Ethical Generative AI

Note: IBM Certifications are optional and are spread across the program duration. IBM content is hosted on the IBM portal but accessible via the course LMS.

Business use cases

UoM-AGAIML Image card Real-World Use Cases

Style Transfer:

Applying GANs for artistic style transfer: using GANs to apply artistic styles to images.

UoM-AGAIML Image card Real-World Use Cases 01

Examining historical Titanic data to understand passenger behaviours and predict survival: 

Creating features to enhance the accuracy of survival predictions. 

UoM-AGAIML Image card Real-World Use Cases 02

How Regression is applied in various industries:  

Using regression models to predict outcomes in different sectors: Applying regression techniques to predict trends and behaviours. 

 

UoM-AGAIML Image card Real-World Use Cases 03

MNIST Digit Classification- Classifying handwritten digits: 

Developing models to accurately identify digits in image datasets. 

 

UoM-AGAIML Image card Real-World Use Cases 04

Forecasting Tourist Footfall: Time series analysis for resource allocation: Using time series models to predict tourist patterns. 

 

UoM-AGAIML Image card Real-World Use Cases 05

Movie Recommendation System: Movie recommendation using collaborative filtering: Building a personalised recommendation system for users.

UoM-AGAIML Image card Real-World Use Cases 06

CI/CD Pipeline for ML Models:

Scaling models using Kubernetes: deploying machine learning models with Kubernetes for scalability.

UoM-AGAIML Image card Real-World Use Cases 07

Predicting Loan Defaults

Predicting loan defaults using ensemble methods: combining models to improve loan default predictions.

UoM-AGAIML Image card Real-World Use Cases 08

Enhancing Patient Care through Text Analytics: Automating clinical note summarisation: Using NLP to summarise clinical notes for healthcare improvements.

Note: The use cases listed above are indicative. Additional use cases may be included or modified based on evolving market trends, and industry requirements.

Hands-on projects

Hands-on projects Sentiment Analysis

Data Cleaning & Visualisation with Git

Hands-on projects Clustering

Retail Banking Customer Churn Prediction

Hands-on projects Image Generation

Predicting House Prices Using Regression

Hands-on projects NLP Models

Employee Attrition Prediction Model

Hands-on projects CNN

Customer Segmentation Using Clustering

Hands-on projects Large Language Models

Time Series Forecasting for Sales

What makes this program so unique accelerated learning

Optimising Deep Learning Models

What makes this program so unique IBM certifications

Sentiment Analysis on Twitter data with Recurrent Neural Networks

What makes this program so unique Global credentials

Advanced Clustering with DBSCAN & GMM

What makes this program so unique job-ready

GANs for High-Resolution Image Generation

What makes this program so unique faculty masterclasses

CNN for Image Classification

University of Melbourne Faculty

UoM - Eduard Photo

Prof. Eduard Hovy

Executive Director, Melbourne Connect | Professor, University of Melbourne | Adjunct Professor, Carnegie Mellon University

Prof. Hovy is the Executive Director of Melbourne Connect, a research and technology transfer centre at the University of Melbourne. He is also a Professor in the School of Co...

UoM Faculty Image - Dr. Jey Han Lau

Dr. Jey Han Lau

Senior Lecturer, School of Computing and Information Systems, University of Melbourne

Academic & Professional Background:
- Ph.D., Graduate Certificate, and Bachelor’s Degree – University of Melbourne
- Senior Lecturer in Natural Language Processing (NLP), Scho...

Career Services Included

Career Services Resume-builder tool

Resume-builder tool

  • 6-month access to DIY resume builder

  • Auto resume creator with optimisation suggestions

  • Unlimited resume iterations within the duration

Career Services Career preparation modules

Career preparation modules

  • Resume and cover letter essentials

  • Maximising LinkedIn and job search strategy

  • Interview preparation and personal branding

Career Services GitHub profile building guidance

GitHub profile building guidance

  • Create your own profile or optimize your existing one.

  • Publish two projects (including the Capstone Project) on your GitHub portfolio.

  • Get insights from domain experts on how/why GitHub is a key differentiator in the interview process.

Certificate awarded on completion of the Advanced Program in AI and Machine Learning with Agentic AI Systems

Certificate awarded on completion of the Advanced Program in AI and Machine Learning with Agentic AI Systems

Participants who achieve a minimum of 70% marks on all mandatory assignments and successfully complete the capstone project will receive a verified certificate of completion from the University of Melbourne. 

FAQs

This program is taught by both University of Melbourne faculty and domain experts. Weekly recorded lectures and weekly live sessions are conducted by domain experts.

Assignments will be graded by industry practitioners who support participants in their learning journeys and/or by the Emeritus grading team.

This program is designed with some of the best faculty to cover relevant topics in a manner that creates positive career outcomes. Career prep modules on resumes, LinkedIn profile optimisation, job navigation and interview preparation are also provided.

We encourage our learners to complete the program to fully understand the concepts and derive valuable learning outcomes. Absent a previous approved request for deferral for the course, learners may request a full refund of all course and miscellaneous fees paid, within fourteen (14) days after course commencement. Application fees for courses are non-refundable and non-transferable. Learners who have previously been granted a course deferral are not eligible for a refund for the course. Partial (or pro-rated) refunds are not offered.

Upon successful completion of the program, you will receive a smart digital certificate. This can be shared with friends, family, schools, or potential employers. You can use it in your cover letter and resume and/or display it on your LinkedIn profile.

You will have access to the online learning platform and all the videos and program materials for 12 months following the program end date. Access to the learning platform is restricted to registered participants per the terms of the agreement.

Yes, the qualifying mark is 70% on all mandatory assignments and successfully complete the capstone project.

Register as soon as possible. Space is limited.

Flexible payment options available.

Starts On