Patrick Prendergast — Senior AI Engineer
AI Agent Infrastructure & MLOps Specialist in Melbourne, Australia
Senior AI engineer with 13 years of experience across cyber security, innovation, and machine learning. Expert in building autonomous agent infrastructure, LLM evaluation pipelines, and MLOps platforms for production-grade AI systems.
Core Expertise
- Agentic Workflows: Designing multi-agent systems and SDLC orchestration.
- MLOps & Engineering: Productionizing models, data pipelines, and evaluation frameworks.
- High-Assurance Systems: Experience in defence, healthcare, and banking sectors.
Professional Experience
- Senior AI Engineer at ROLLER (Melbourne). Leading development of ROLLER iQ, focusing on agent infrastructure, automated evaluation pipelines, and scalable MLOps platforms.
- Machine Learning Engineer at ANZ Bank. Engineered production data pipelines for fraud detection, processing over 25M daily transactions with high availability.
- Research Engineer at University of Melbourne. Developed deep learning models for medical imaging (MS lesion detection) with >95% AUC.
- Cyber Security Researcher at Defence Science & Technology Group. Focused on high-assurance software and secure systems engineering.
Open Source AI Projects
- Chief Wiggum — Agentic SDLC orchestration platform. Orchestrates multiple AI models (Claude, GPT, Gemini) for test-driven development and code review.
- Nifti.NET — High-performance .NET library for medical imaging (NIfTI), with 13K+ downloads on NuGet.
- Hue MCP — Implementation of the Model Context Protocol (MCP) for smart home automation.
- AFL Prediction — Machine learning ensemble for match forecasting using gradient boosting and logistic calibration.
Links
Contact
pat@plwp.net
Visualisation inspired by Lai et al. (2018) — Next generation histology methods for three-dimensional imaging of fresh and archival human brain tissues, Nature Communications.