I’m Alan Huynh, an ML and software engineer based in Australia, focused on building production-grade AI systems—from computer vision and time-series forecasting to physics-informed neural networks and geospatial analytics. Over the past few years I’ve worked across energy, environmental modelling, and satellite/remote-sensing applications, often combining rigorous ML experimentation with practical engineering for deployment on real infrastructure. Recently I’ve been building AI products and platforms end-to-end, including an AI subtitling system with multi-agent quality control and human-in-the-loop review, and an automated AEMO NEM trading system using mathematical optimisation. I also enjoy bridging research and engineering: for example, adapting vision models for robust real-world performance, optimizing inference for edge devices (including TensorRT/Jetson), and shipping reliable pipelines with Docker, AWS, CI/CD, and automated testing.

Alan Huynh

I’m Alan Huynh, an ML and software engineer based in Australia, focused on building production-grade AI systems—from computer vision and time-series forecasting to physics-informed neural networks and geospatial analytics. Over the past few years I’ve worked across energy, environmental modelling, and satellite/remote-sensing applications, often combining rigorous ML experimentation with practical engineering for deployment on real infrastructure. Recently I’ve been building AI products and platforms end-to-end, including an AI subtitling system with multi-agent quality control and human-in-the-loop review, and an automated AEMO NEM trading system using mathematical optimisation. I also enjoy bridging research and engineering: for example, adapting vision models for robust real-world performance, optimizing inference for edge devices (including TensorRT/Jetson), and shipping reliable pipelines with Docker, AWS, CI/CD, and automated testing.

Available to hire

I’m Alan Huynh, an ML and software engineer based in Australia, focused on building production-grade AI systems—from computer vision and time-series forecasting to physics-informed neural networks and geospatial analytics. Over the past few years I’ve worked across energy, environmental modelling, and satellite/remote-sensing applications, often combining rigorous ML experimentation with practical engineering for deployment on real infrastructure.

Recently I’ve been building AI products and platforms end-to-end, including an AI subtitling system with multi-agent quality control and human-in-the-loop review, and an automated AEMO NEM trading system using mathematical optimisation. I also enjoy bridging research and engineering: for example, adapting vision models for robust real-world performance, optimizing inference for edge devices (including TensorRT/Jetson), and shipping reliable pipelines with Docker, AWS, CI/CD, and automated testing.

See more

Experience Level

Expert
Expert
Expert
Expert
Expert
Expert
Expert
Intermediate
See more

Language

English
Advanced

Work Experience

Machine Learning Consultant at ProtoFlow Pty Ltd
June 1, 2025 - Present
Built an AI subtitling platform processing 1,000+ hours of video across 10 languages using an automated multi-agent MQM quality pipeline across Claude/Gemini/GPT, ASR via Deepgram, and human-in-the-loop review. Authored 20+ LLM training problems designed to defeat strong baselines. Built an automated AEMO NEM trading system using mathematical optimisation for dispatch scheduling, achieving a 29% revenue uplift over the production model. Led design and implementation of a proprietary POS and kiosk platform for a multi-site F&B operator (1000+ customers/day average), forming a three-person operations team and owning delivery end-to-end. Built a Moodle LLM assistant for an education provider serving 200 students daily.
Machine Learning Engineer at Fair Dinkum Systems
December 1, 2024 - February 28, 2025
Built dual-stream Vision Transformers in PyTorch for battery remaining-useful-life prediction and evaluated across six datasets including the Stanford fast-charging dataset. Developed a time-series feature pipeline using wavelet transforms, spectrograms, and Fourier analysis, and added contrastive learning to improve out-of-domain robustness. Worked in a 4-person team with pytest and GitHub Actions.
Software Engineer at Math Song Pty Ltd
September 1, 2024 - September 1, 2025
Led full-stack Django platform development to automate manual business operations. Implemented payment integrations supporting $200K+ per month transaction volume with no payment-path incidents during tenure. Managed GCP infrastructure using Docker, Terraform, and CI/CD pipelines including VM provisioning.
Machine Learning Engineer at AIgorithm
February 1, 2024 - September 30, 2024
Delivered satellite-derived emissions metrics at Australian ABS Mesh Block resolution (368k units) for a carbon initiative serving banking and retail clients. Built semi-supervised computer vision on multi-spectral imagery, including self-supervised pretraining to adapt European land-classification frameworks to Australian conditions. Implemented an LSTM-to-XGBoost pipeline estimating net carbon exchange and methane from NASA Earthdata inputs over ALUM land-use classes, containerised for AWS ECR and Fargate. Built production ETL for multi-spectral imagery on AWS Glue (Spark), EMR, and Lambda.
Data Scientist at Lux Aerobot
January 1, 2022 - January 31, 2024
Developed computer vision for wildfire detection from multi-spectral imagery using stratospheric balloon platforms supporting environmental monitoring and disaster recovery in Canada and Australia. Optimised models for NVIDIA Jetson quantisation and TensorRT to reduce model size by 60% for real-time field inference. Designed the full gondola stack (hardware, firmware, and deployment software) and built AWS training pipelines on SageMaker, Neo, and Greengrass.

Education

Bachelor of Information Technology at Melbourne Polytechnic
February 1, 2024 - October 1, 2026
Accelerated Program at Deakin University
February 1, 2023 - October 1, 2023

Qualifications

AWS Certified Machine Learning Specialty (MLS-C01)
March 1, 2025 - March 1, 2028

Industry Experience

Energy & Utilities, Financial Services, Education, Media & Entertainment, Computers & Electronics, Software & Internet, Professional Services