This internship position is being offered through the Computing Internships course COMP4820/8830. You must follow the instructions on the Computing Internships webpage to apply.
Organisation:
Shine has a deeply technical culture. We love learning and sharing our experiences with each other, our clients, and software communities.
Project:
Machine Learning Training Platform
Background Shine Solutions aims to develop an end-to-end solution to set up and deploy an AWS environment with appropriate tools and configurations for training machine learning models using customer-provided datasets. The goal is to create a scalable, automated, and user-friendly platform that enables efficient model training, fine-tuning, and deployment. The project will leverage AWS-native services to ensure high performance, cost efficiency, and ease of integration with existing cloud solutions. The platform will also include an intuitive interface for data ingestion, processing, and storage, allowing customers to seamlessly upload datasets for training purposes.
Business Requirements
The AI/ML team at Shine Solutions has outlined the following core features and design requirements for the solution:
Design:
The solution must be hosted and operated on AWS cloud services. Building ML models is an iterative process that requires a large amount of experimentation. Therefore, the solution must include a system which ensures the model building is reproducible, repeatable and auditable. It should use a modular design approach with microservices and serverless functions where feasible. The components should adhere to loose coupling principles to ensure flexibility and scalability.
Features:
A customer-facing interface for data ingestion and storage. The ability to train machine learning models leveraging AWS services, such as SageMaker A repeatable process for pre-processing, training and evaluation of the ML models. A mechanism to fine-tune machine learning model training. Storage for model outputs and trained models. A mechanism to store, share, and distribute trained models to a wider audience in an organisation. A testing framework to evaluate the accuracy of trained models. The ability to utilise existing pre-trained models to assist with training or preprocessing input data. A CI/CD pipeline for testing and deployment of the solution.
Technical Requirements Deployment:
Use of AWS-native services, including but not limited to Amazon S3, AWS Lambda, Amazon SageMaker, and AWS CloudFormation. Implementation of a CI/CD pipeline using AWS CodePipeline and AWS CodeBuild for automated testing and deployment. Testing: Unit tests for the API and serverless components of the solution. Unit tests for the data ingestion and training pipeline. Model validation and evaluation to ensure accuracy before deployment.
PROJECT DELIVERABLES
Source code saved in Shine Solution’s private Git repository. Fully functional AWS-based machine learning training platform. Documentation for setup, deployment, and usage. Presentation on the project architecture and implementation details.
Required/Preferred Technical Skills:
Knowledge of AWS cloud services or a willingness to learn. Experience with AWS serverless technologies. Proficiency in Python, Typescript, or Javascript. Experience with CI/CD pipeline implementation. Understanding of machine learning workflows.
Required/Preferred Professional Skills:
Teamwork, Presentation Communications & Story-telling, Problem solving, Analytical.
Special Requirements:
The intern must be an Australian Citizen.
Type of Internship:
Hybrid internship (remote or in-person).
The internship will be paid. The intern/s will be engaged as a casual employee
How to Apply:
Please follow the instructions on the Computing Internships webpage.