Jesse Ashton Email & Phone Number

Auckland, Nz

Auckland, New Zealand

* Automated customized pre-sales reports for prospective customers using the ChatGPT API to scale data quality comparisons against competitors (Pandas, Plotly, Jupyter Notebooks). Achieved a 100% conversion rate to the active evaluation phase of the sales cycle, while reducing associated costs by 50% and increasing efficiency by 38%.* Implemented data observability tests (Soda, Snowflake) and accompanying snapshots (dbt) run automatically through Prefect flows to maintain data quality on tens of thousands of events ingested per day.* Developed a custom internal UI (AJAX, jQuery, HTML) including implementing new API endpoints (Ray API / Python) which enabled non-technical colleagues to modify and run predictive models for customers. This eliminated a major bottleneck, tripling throughput for customer support requests. * Developed ETL processes (Elasticsearch, dbt, Snowflake) and a dashboard (Metabase) for the sales team to monitor API usage across customer base, identifying those at high risk of churn. * Created SQL (Snowflake) and Python scripts to reuse GeoJSON polygons from over 5,000 past events, leading to savings of tens of thousands per annum in contractor fees.* Developed Python scripts using duckdb to efficiently detect missing data from large customer datashares (s3) which helped maintain contracts with two customers. * Evaluated and enhanced a quantile regression model for predicting regions of interest for events impacting customer locations, including developing new data sources (loaded into Redshift), feature engineering (transformation and normalisation) and research for optimal parameterisation, improving performance by 150%* Evaluated a model to predict consumer spending associated with events leading to a 5x improvement in model accuracy. Involved visualising data (Pandas, Seaborn, Plotly), documenting findings (Confluence) and delivering knowledge sharing workshops. Products built from this model helped close new customer contracts.

Auckland, New Zealand

•Trained machine learning algorithm (random forest classifier in ITK-SNAP) to detect cells in microscopy images.•Used RStudio (including dplyr package, very similar to SQL) to compare measurements of brain cell anatomy and electrical activity between normal and disease states (t test, analysis of variance, and non-parametric Wilcoxon signed-rank test). Spatial data mapped in Tableau. Study published in peer-reviewed journal: doi.org/10.1152/ajpheart.00020.2020•Developed MATLAB graphical user interface for analysis and classification of time-varying physiological signals with automatic event detection using cross-correlation based template matching.•First author on 4 original research articles and reviews in peer-reviewed journals (google scholar).•Co-supervised and coached postgraduate students (1x PhD, 1x Masters) which involved working closely in the lab teaching experimental and data analysis skills

Auckland, New Zealand

•Helped develop end-to-end platform for visualisation of rainwater and wastewater monitoring data and interactive simulation of water supply for regional councils.•Responsible for front and back end development, database schema design and maintenance, writing SQL queries and stored procedures, and demonstrating software for clients.•Functionality I developed included: -The main interactive plotting interface using the Google Charts API.-A microservice which updated the SQL database when new monitoring data was received through the telemetry network.-Online generation of interactive 3D pipe models using Google SketchUp (scripted in Ruby). -Backend interface for command-line execution of external application for water supply simulation.-A tablet application for technicians to record wastewater monitoring data in the field. •Technologies used: C#.NET, JavaScript, Google Maps API, MySQL, PostgreSQL, MS Access.

Auckland, New Zealand

•Designed new computational model for predicting deformation induced stress in heart muscle and performed optimization to fit model to experimental data. •Co-wrote 1 research article.•Technologies used: Wolfram Mathematica

Doctor Of Philosophy (Phd), Biomedical Engineering

Activities and Societies: Engineers Without Borders New Zealand •Awarded my PhD with no corrections required to my thesis (rarely achieved.

Master Of Engineering (Me) With First Class Honours, Biomedical Engineering

•GPA: 9 (A+) •Developed realistic finite element models of heart muscle using image processing (LabVIEW) to extract anatomy.

Bachelor Of Engineering (Be) With First Class Honours, Biomedical Engineering