Conservation · Audio ML01
Audio Signatures: Bioacoustics for Wildlife ID
Offline acoustic identification across 234 Pantanal taxa, from sound-event detection to an ONNX inference pipeline.
Independent AI projects · Professional leadership · Sydney
Exploring practical AI, one project at a time.
My professional experience spans more than 25 years across technology, product engineering, delivery and commercial leadership in APAC and global markets. The AI projects shown here are independent passion projects I pursue from personal interest—not client engagements or professional work. They give me a hands-on way to explore machine learning across conservation, decision intelligence, generative systems, finance and structured prediction.
25+Years in technology and business leadership
07Independent personal projects across AI problem areas
09Selected published resources and tools
Selected work
Research that connects models to real problems.
These are independent passion projects pursued outside my professional role—not work delivered for clients or employers. Each page shares the approach, evidence and practical limitations behind the exploration.
Conservation · Computer vision02
Computer Vision for Wildlife ID
Jaguar re-identification using frozen DINOv2 features and metric learning, achieving 0.950 private mAP on scarce data.
Decision intelligence · RAG03
Decision Lens
Evidence-grounded recommendations that combine retrieval, behavioral science and traceable source attribution.
Generative audio · Privacy04
SonarAI Voice Synthesis
A local-first speech system with multilingual voices, zero-shot cloning, SSML control and broadcast-ready output.
Quantitative finance05
Stocks Value Investor
A rules-driven equity research screener with sector-aware valuation logic and LLM-assisted thesis validation.
Computer vision · Identity06
Real-Time Face Recognition
A production-oriented identity-verification pipeline designed around responsive WebRTC streaming and modern embeddings.
Tabular ML · Customer ranking07
Gradient-Boosted Decision Trees
Use customer data to identify who is most likely to buy or leave, prioritise audiences and focus limited outreach resources. Learn how AI-powered ranking can inform campaign targeting—and why controlled tests are still needed to measure the extra results a campaign delivers.
Continuous practice
Selected credentials.
Ongoing learning across AI infrastructure, cloud, model development and responsible deployment supports a practical, research-led approach.
Google Certified — Generative AI Leader
Adobe Qualified — Adobe Experience Platform
NVIDIA-Certified Associate — AI Infrastructure and Operations
Microsoft Certified — Azure Fundamentals
Statistical Learning with Python
Oracle Certified Professional — Java
Attention Mechanism certification
Google Cloud Feature Engineering
Professional Machine Learning Engineer
MLOps for Generative AI
Machine Learning Operations with Vertex AI — Model Evaluation
Responsible AI for Developers — Fairness and Bias
Model Armor — Securing AI Deployments
Connect
Bring strategy closer to the technology.
For conversations on enterprise AI, applied machine learning and responsible deployment, connect with me on LinkedIn.