Lecturer & Unit Coordinator · Melbourne Institute of Technology
Founder of AcadGate and Revlyzer

Osamah A. Mahdi, PhD

AI and machine learning academic, researcher and product builder. I teach machine learning and data analytics, research how models stay reliable when data changes over time, and build AI products on large language models and retrieval-augmented generation.

Get in touch Google Scholar LinkedIn
Publications
35
Citations
428
h-index
13
Student evaluation
4.6/5
Years teaching
8+
Portrait of Osamah A. Mahdi in a navy suit
Melbourne, Australia

01 · About

Teaching, research and building, in one practice

I hold a PhD in Computer Science and Information Technology from La Trobe University and have more than eight years of teaching at five Australian institutions. At the Melbourne Institute of Technology I design, coordinate and teach machine learning, data science and big data analytics units at undergraduate and postgraduate level, face-to-face, online and blended. My student evaluations average 4.6/5, and I received MIT’s Teaching Excellence Award in 2024.

Alongside teaching, I design and build AI applications using large language models and retrieval-augmented generation, including AcadGate and Revlyzer. My research has produced 35 publications (h-index 13), including articles in Knowledge-Based Systems and IEEE Access, and I supervise eight Master by Research students.

02 · Hands-on AI product development

Products I have designed and built

End-to-end ownership, from architecture and AI pipelines to security, privacy and working with customers.

AcadGate

Founder · Since 2025

AI teaching and assessment platform for universities

A RAG pipeline draws on a lecturer’s own course documents to generate question banks, answer student questions with citations, mark work against the lecturer’s rubric and report on engagement. Nothing reaches students until the lecturer approves it. Connects to Moodle, Canvas and Blackboard through LTI 1.3 with single sign-on.

FastAPI PostgreSQL React Google Cloud Run LTI 1.3
acadgate.com

Revlyzer

Founder · Since 2025

AI review analytics for local businesses

Turns Google reviews into practical insight. An LLM (Anthropic’s Claude) analyses sentiment, recurring topics and urgent complaints, then produces a priority action plan, competitor SWOT benchmarking, multi-location dashboards and PDF, Excel and Word reports. Paying customers include a multi-hospital healthcare group.

Claude LLM React TypeScript
revlyzer.com

TrusTurn

Founder · Since 2026

Rotating savings (ROSCA) fintech platform In development

Lets small groups pool and rotate contributions securely: a double-entry ledger, Stripe Connect payments and payouts, KYC checks, automated contribution cycles and native iOS and Android apps. Fully working in payment test mode, with AFSL and AUSTRAC compliance under way ahead of an Australian launch.

Stripe Connect iOS & Android KYC
trusturn.com

Also developed DevWise.AI, a conversational assistant for novice developers.

03 · Teaching

Academic and teaching appointments

  1. Melbourne Institute of Technology

    Aug 2021 – present

    Lecturer and Unit Coordinator · Student evaluation 4.6/5

    Coordinate and teach across the Master and Bachelor of Data Analytics and the Master and Bachelor of Networking: Predictive Data Analytics, Software Practice for Big Data Analytics, Data Science, Mathematical and Statistical Methods, ICT Practices, Research Methods and Project Design, Data Science Fundamentals, Network Management in Organisations, Capstone Project, Platform Technologies and Operating Systems. Design learning materials and practice-based assessment, and lead and mentor sessional staff.

  2. Torrens University Australia

    Jul 2024 – Aug 2025

    Sessional Academic / Teaching Assistant · 93–100% satisfaction

    Introduction to Cloud Computing using AWS, Enterprise Computing Systems, Research Methodologies and Introduction to Programming.

  3. La Trobe University

    Jul 2017 – Nov 2021

    Sessional Tutor / Subject Support / Lab Demonstrator · 3.95–4.50/5

    Machine Learning, Data Exploration and Analysis, Data Warehouse Concepts and Design, Database Fundamentals and Systems Management, Information Systems, Object-Oriented Programming and Entrepreneurship in IT. Designed assessments delivered concurrently across Melbourne, Bendigo and Sydney campuses, recruited lab tutors and chaired weekly teaching-team meetings.

  4. RMIT University

    Mar 2021 – Nov 2022

    Sessional Teaching Assistant · Mean 4.47/5 (2021)

    Algorithms and Analysis, in Java and Python, across four semesters.

  5. Monash University

    Jul 2021 – Nov 2021

    Sessional Teaching Assistant · 93/100 (Very High)

    Programming Fundamentals in Java, against a faculty average of 82/100.

  6. Didasko Learning Resources

    Jul 2019 – Mar 2022

    Moderator

    Information Systems and Database Fundamentals on the Cloud.

04 · Research

Machine learning for data that changes

My research develops methods for detecting and adapting to concept drift, and applies them to cybersecurity, anomaly detection and other real-world systems. I am extending this work to LLM-based systems: how retrieval-augmented and other LLM applications stay reliable as the documents and data behind them change.

Algorithms and techniques

LLMs and RAG, deep learning, explainable and multimodal AI, concept drift detection and adaptation, data stream mining, ensemble and adaptive learning, federated and edge learning, robust and adversarial ML.

Cybersecurity and anomaly detection

ML-based intrusion detection, real-time anomaly detection and false-alarm reduction, adversarial defence, privacy-preserving and encrypted-traffic analysis, behaviour-based risk analysis.

Applications

IoT and streaming data, cloud-edge systems, critical infrastructure and energy, healthcare, enterprise information systems, and AI in education.

Research grant: Diversity Measure for IoT Intrusion Detection Systems, Melbourne Institute of Technology, 2023–2024 (Investigator).

05 · Publications

Selected publications

Full list on Google Scholar

Journal articles and book chapter

  1. 2025

    Federated learning under concept drift: A systematic survey of foundations, innovations, and future research directions

    Mahdi, O. A., Pardede, E., Bevinakoppa, S., & Ali, N. · Electronics, 14(22), 4480 · Q2

  2. 2025

    Unveiling the role of diversity measure in concept drift detection: A comparative analysis and future prospects

    Mahdi, O. A., Ali, N., Pardede, E., & Das, B. · Book chapter in Enhancing Data-Driven Electronics through IoT, IGI Global, pp. 1–32

  3. 2024

    Roadmap of concept drift adaptation in data stream mining, years later

    Mahdi, O. A., Ali, N., Pardede, E., Alazab, A., Al-Quraishi, T., & Das, B. · IEEE Access, 12, 21129–21146 · Q1, IF 3.9

  4. 2023

    Routing attacks detection in 6LoWPAN-based Internet of Things

    Alazab, A., Khraisat, A., Singh, S., Bevinakoppa, S., & Mahdi, O. A. · Electronics, 12(6), 1320 · Q2, IF 2.69

  5. 2021

    A hybrid block-based ensemble framework for the multi-class problem to react to different types of drifts

    Mahdi, O. A., Pardede, E., & Ali, N. · Cluster Computing, 24(3), 2327–2340 · Q2, IF 2.6

  6. 2020

    Diversity measure as a new drift detection method in data streaming

    Mahdi, O. A., Pardede, E., Ali, N., & Cao, J. · Knowledge-Based Systems, 191, 105227 · Q1, IF 8.8

  7. 2020

    Fast reaction to sudden concept drift in the absence of class labels

    Mahdi, O. A., Pardede, E., Ali, N., & Cao, J. · Applied Sciences, 10(2), 606 · Q2, IF 2.84

  8. 2018

    Exploring instances for matching heterogeneous database schemas utilising Google similarity and regular expression

    Mahdi, O. A., Ibrahim, H., Affendey, S. L., Pardede, E., & Cao, J. · Computer Science and Information Systems, 15(2), 295–320 · Q2

  9. 2017

    An approach for instance-based schema matching with Google similarity and regular expression

    Mahdi, O. A., Ibrahim, H., & Affendey, L. S. · International Arab Journal of Information Technology, 14(5), 755–763 · Q2

Selected conference papers

  1. 2026

    Federated drift-aware learning: Current challenges and future directions

    Mahdi, O., Bevinakoppa, S., & Ali, N. · International Conference on Advances in Computing Research, Springer, pp. 222–231

  2. 2026

    Text anomaly detection with LLM embeddings: A review

    Awal, M. A., Mahdi, O. A., & Rahimi, I. · MIT ICETE Conference, Sydney · doi:10.25397/0m50-ge79

  3. 2026

    Enhanced IoT intrusion detection: A hybrid framework integrating decision tree and one-class SVM

    Singh, S., Mahdi, O. A., Bevinakoppa, S., & Alazab, A. · MIT ICETE Conference, Sydney · doi:10.25397/n0br-hp82

  4. 2026

    Towards explainable multi-modal AI for Alzheimer’s diagnosis: A comparative review of vision transformers, Grad-CAM, and language models

    Aryal, S., Al-Jumaily, A., & Mahdi, O. A. · MIT ICETE Conference, Melbourne and Sydney · doi:10.25397/v7xe-8q50

  5. 2024

    Online concept drift detector: Optimally balancing delay detection, runtime, memory, and accuracy

    Mahdi, O. A., Ali, N., Pardede, E., & Al-Quraishi, T. · Procedia Computer Science, 237, 559–567

  6. 2024

    Diversity measure for concept drift detection in data streams · Multi-criteria analysis of concept drift detection algorithms

    Mahdi, O., Bevinakoppa, S., & Singh, S. · ICETAS 2024, 9th International Conference on Engineering Technologies and Applied Sciences, IEEE

  7. 2023

    Enhancing IoT intrusion detection system performance with the diversity measure as a novel drift detection method

    Mahdi, O. A., Alazab, A., Bevinakoppa, S., Ali, N., & Khraisat, A. · 9th International Conference on Information Technology Trends (ITT), IEEE, pp. 50–54

  8. 2021

    KAPPA as drift detector in data stream mining

    Mahdi, O. A., Pardede, E., & Ali, N. · Procedia Computer Science, 184, 314–321

  9. 2018

    Combination of information entropy and ensemble classification for detecting concept drift in data stream

    Mahdi, O. A., Pardede, E., & Cao, J. · Australasian Computer Science Week Multiconference, pp. 1–5

06 · Education

Degrees

2016 – 2021

PhD, Computer Science and Information Technology

La Trobe University, Melbourne

Thesis: Diversity Measures as a New Concept Drift Detection Method in Data Stream Mining.

2011 – 2014

Master of Research, Computer Science

Universiti Putra Malaysia

Thesis: A New Approach for Instance-Based Schema Matching.

2005 – 2009

Bachelor of Science, Computer Science

University of Babylon, Iraq

Certification

Google Data Analytics Professional Certificate

07 · Skills and service

Toolkit

Machine learning and AI

PyTorchTensorFlowscikit-learnWekaMOALLM apps with ClaudePrompt engineeringRAG pipelines

Programming and web

PythonRJavaC/C++TypeScriptReactNode.jsFastAPI

Data, cloud and analytics

AWSGoogle Cloud RunSupabasePostgreSQLOracleSQL ServerTableauPower BI

Professional service

Peer reviewer for Knowledge and Information Systems, The Journal of Supercomputing, Machine Learning, Applied Artificial Intelligence, MDPI Applied Sciences and Electronics, and others. Member of IEEE (Computational Intelligence Society), ACM and the Australian Computer Society.