Tom

About Me

As an AI/ML Cloud Support Engineer at Amazon Web Services, I help enterprise customers design and deploy production machine learning infrastructure on AWS, prioritising scalability, availability, cost optimisation, and general performance. I hold a Bachelor of Engineering (Honours) in Engineering Science and a Bachelor of Commerce in Finance from The University of Auckland.

Technical Skills

SageMaker, Bedrock, MLflow, Airflow, VPC, DynamoDB, S3, Kinesis, CloudWatch, EKS, CloudFormation, Terraform, Python, SQL, Docker, PyTorch, Scikit-Learn

Interests

Surfing, Cycling, Swimming, Running, Skiing, Hiking, Hawke's Bay Marathon 2025, Auckland Marathon 2023

Publications

Two-hourly prediction of hypoxic-ischemic injury timing using EEG during the latent phase in near-term fetal sheep

Purkis, T., Davidson, J., Gunn, A. J., Maurice Loomes, C., Bennet, L., & Abbasi, H. (2025)

2025 47th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

View Publication →

Transformers predict hypoxia-ischemia timing in term fetal sheep EEG in the key 2-Hour window post-insult

Matthews, A. R., Davidson, J., Purkis, T., Gunn, A. J., Bennet, L., & Abbasi, H. (2025)

2025 47th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

View Publication →

Work Experience

AI/ML Cloud Support Engineer

Amazon Web Services

Dec. 2025 – Present

Guide the design and deployment of end-to-end customer ML pipelines, prioritising cost optimisation and scalability.
Collaborate with service teams to relay customer infrastructure needs and influence ML platform offerings.
Developed and pitched an ML lead generation PoC to senior leadership with the potential to produce ~30,000 verified leads per year for account teams.

AI/ML Cloud Support Associate

Amazon Web Services

Feb. 2025 – Nov. 2025

Provide integration and implementation guidance to large global customers deploying production ML pipelines on AWS.
Diagnose and resolve critical production issues for enterprise customers.

Summer Research Scholarship

Auckland Bioengineering Institute

Nov. 2024 – Feb. 2025

Researched the novel application of ML to hypoxic-ischemic encephalopathy timing prediction, achieving AUC of 0.94 ± 0.05 in classification.
Resulted in two ML research papers for the EMBC IEEE 2025 Conference.

Machine Learning Engineer

Presently

Mar. 2024 – Nov. 2024

Developed an event-driven, decoupled computer vision system on AWS that processes crowd reactions to conference presentations, analysing sentiment and engagement.

Investment Banking Analyst

Cranleigh Partners Ltd

Sep. 2022 – Nov. 2023

Financial modelling and strategic analysis for Cranleigh and its clients.

Founder and Director

Pascal Group Ltd

Dec. 2021 – Mar. 2024

Web development studio providing services to SMEs in Auckland.
Sample site: mobilemedic.co.nz.

Projects

Automated MCP Server Generator

Built tooling to automatically convert OpenAPI schemas into MCP servers, enabling seamless AI agent integration.

Python OpenAPI MCP

Learn More →

MCP Prompt Management Platform

Developed and deployed a cloud-native prompt management system with remote hosting on CloudFlare.

CloudFlare MCP

VR Treatment for TBI

Honours Project: VR as a treatment for Traumatic Brain Injury using multimodal ML techniques.

VR ML Healthcare

Awards & Certifications

Movember Mo Your Own Way Award 2025

Running around Rangitoto for 24 hours

Learn More →

AWS Solutions Architect Associate

Design of cost and performance optimized solutions on AWS

Amazon Mentor Accreditation Program

Internal Amazon leadership program

Dean's Honours List 2020 & 2021

Awarded to the top 5% of cohort or for attaining a GPA above 8.25

Certificate of Distinction

STATS 370 course at The University of Auckland

Get in Touch

Always happy to connect with fellow engineers to discuss interesting problems in ML and cloud infrastructure.

Auckland, New Zealand

Hi, I'm Tom Purkis