About
I currently work as an Analytics Engineer at Airtasker. I have particular
interest in CV and MLOps.
I am very interested in the applications of my field in agriculture and what these applications can do to
strengthen Australian industry.
- Website: www.cooperjwhite.com
- City: Wollongong, NSW
- Degree: B.S. Mechatronic Engineering
- Email: cooperjwhite0@gmail.com
Resume
Summary
Cooper White
Reliable and deadline-driven Data Consultant with 2 and a half years experience in Data Analytics, Data Engineering and ML
- Sydney, NSW
- cooperjwhite0@gmail.com
Education
B.S. Mechatronic Engineering (Honors)
2017 - 2020
University of Wollongong, Wollongong, NSW
Professional Experience
Analytics Engineer
Jan 2024 - Present
Airtasker, Sydney, NSW
Data Consultant
Aug 2022 - Jan 2024
Servian, Sydney, NSW
- Migrating a supply-chain companies existing data pipelines from Jenkins/Informatica to GCP Composer, Cloud SQL and Cloud Run.
- Responsible for developing custom deep learning models with Tensorflow and serving them on AI Platform for use in Google Earth Engine. This solution allowed Servian to become one of only two recognised APAC partners in GEE and the only partner globally to present an ML model using this custom methodology.
- Utilizing SQL and SAS in a financial institutions Remediation team to explore and analyse large scale data warehouses in order to identify impacted customers and derive the remediation value.
Process Analytics Engineer
Aug 2021 - Jul 2022
MM Kembla, Wollongong, NSW
- Exploring custom deep learnimng anomaly detection models for fault identification.
- Uses machine learning to identify key production parameters for optimisation.
- Developed web application for automated data analysis and report generation.
- Developed ETL pipelines using Python and SQL on Linux Servers for use in PowerBI dashboards.
Mechatronic Engineer/Production Supervisor
Dec 2020 - Jun 2021
Koenigs Isolators and Enclosures, Wollongong, NSW
- Utilised AutoCAD and Inventor to develop general arrangements as well as watercutter and bending programs.
- Developed stock and job management systems for optimizing throughput in a fast paced manufacturing environment.
- Collaborating with workshop tradesmen and engineers to implement quality and reliability protocols and strategies.
Projects
CDP ETL Pipeline Migration
WooliesX
Python, SQL, Postgres, GCP, Cloud Run, Cloud Composer
Company was moving their existing Informatica/AWS pipelines to GCP. This migration involved analysing legacy Bash and/or Informatica pipeline architecture in order to recreate their behaviour in a Python script executed on Cloud Run, as well as developing a corresponding DAG to be used for orchestration of the pipeline with Cloud Composer.
Web App for tracking consultant onboarding status
Servian
React, Javascript, Python, NoSQL, API Gateway, Cloud Run, Firestore
Identified lack of a single source of truth for the status of consultants onboarding to a client, I worked with internal stakeholders to identify points of issue in current process and to map out requirements. Solution was to create an interactive dashboard using a React frontend that utilized a microservice based backend using Cloud Run exposed via API Gateway and a NoSQL Firestore database.
Terrain Cost Mapping for Seismic Testing Vehicle
Servian
Javascript, Google Earth Engine
Working on a pre-sale proof-of-concept for a client where we collated multiple GIS datasets and produced an interactive GEE app that allowed the client to select an area of interest and assign weights to different environmental and geological indexes such as NVDI and Terrain Slope, the app would then produce a "cost map" of the terrain to assist the client in trajectory planning for their vehicles.
Real-time carpark occupancy analytic dashboard
Personal
Python, GCP, Pub/Sub, Dataflow, Bigquery, Looker
Utilized TfNSW’s APIs to create a streaming pipeline utilizing Cloud Functions, Pub/Sub and Dataflow to stream data into Bigquery. Exposed data to a Looker dashboard to allow for monitoring and analytics of carparks located across NSW.
Book Recommender System
Personal
Python, Pytorch, DBT, GCP, Vertex AI
Utilized DBT with Goodreads and Amazons public review datasets to create a feature-engineered dataset for training a Recommendation System. I trained different recommender models such as KNN, Matrix Factorization and LSTM recommender systems with PyTorch. Exposed model via Vertex AI.
Bushfire Prediction Model/Google Earth Engine Go-To-Market
Servian
Python, Tensorflow, Google Earth Engine
Using Google Earth Engine for both data collection and inference, I trained a custom deep learning model to estimate bushfire risk of an area. Trained the model using Tensorflow and served via AI Platform. I documented and presented my solution to Google for Servian’s GEE partner interview, which allowed us to become the only partner globally to present a working model. I also played a key role in developing the go-to-market material and pre-sale consultation with clients.
Client Remediation Data Analyst
Westpac
SQL, SAS
Utilizing SQL and SAS, identify impacted customers eligible for remediation and calculate their fiscal entitlement while adhering to ASIC commitments and regulations. Required deep knowledge of the product databases and efficient data management. I was responsible for communicating with business stakeholders to identify and translate key business rules into technical requirements and procedures.
Anomaly Detection of Wall Thickness
MM Kembla
Python, SQL, Tensorflow
Developed an anomaly detection system using autoencoders in Tensorflow to flag potential defective tube stock. Data was collected from four different wall thickness measuring devices, and was processed using an ETL pipeline written in Python and SQL. Data was then displayed in a PowerBI dashboard.
Producton Analytics Report Platform
MM Kembla
Python, SQL
Developed an interactive web platform that allowed users to generate custom reports for standard production metrics, also allowed for emailing reports to others ad-hoc or on a recurring basis.