Iris Meredith

Scientific Programmer

Iris joined Bodeker Scientific in November 2021 on a 6-month Callaghan Innovation R&D Career grant. Iris is a scholar and engineer based in Hamilton, New Zealand. She extensive postgraduate study experience across multiple fields, and has fitted an unparalleled amount of work experience in her six years in the workforce. Iris is a strong developed across multiple languages, and has developed multiple internal tools to solve problems in data analysis, visualisation and mathematics. She has five years experience in machine learning, and executed machine learning projects in fields ranging from computer vision to data synthesis - thus gaining a strong understanding of machine learning techniques and how to apply them practically.

Academic qualifications

  • November 2020 - November 2021: Masters of Engineering, University of Auckland

  • March 2012 - May 2017: Bachelor of Engineering (hons), University of Auckland

Positions held

  • August 2021 - Present: Manager and Web Designer, Instruments of Mayhem

  • November 2020 - Present: Research Assistant, University of Auckland

    • Developed a pipeline for synthetic data generation using both Bayesian Network methods and Generative Adversarial Networks, allowing a large volume of previously inaccessible transport data to be used for machine learning applications.

  • December 2018 - November 2020: Junior Data Scientist, Kantar NZ

    • Led development on a product for combining and visualising multiple geographic datasets and making them accessible to clients through a web-based interface. The product has been sold to a client and is currently being offered as a consulting tool: launch of the web interface is planned for early next year.

    • Performed the analysis and wrote the technical notes for Metro Magazine’s Best Schools issue in 2019, which is read by tens of thousands of Aucklanders each year.

    • Performed Data Science projects for a number of clients, with a very high rate of client satisfaction.

  • April 2017 - December 2018: Research Assistant, LIC

    • Helped develop a machine learning algorithm to assess pasture cover on dairy farms and verified the results statistically. This has now gone to market as the LIC SPACE service and the team were finalists in the 2018 KuDos Agricultural Science awards.

    • Worked on a project aiming to detect lameness in dairy cattle using image processing methods (Convolutional Neural Network + point tracking between frames). Initial results are promising.

    • Initiated a project to leverage LIC’s customer data for analytics applications. This project is ongoing.

  • February 2016 - December 2016: Software Developer, Marshall Day Acoustics

    • Developed a Finite Element solver and graphical interface for use in finding the resonant modes of small rooms such as recording studios. The solver saw use in at least two design projects as of when I left Marshall Day.

    • Used an analytic model of sound transmission through double - plate plaster walls to calculate attenuation coefficients for partition walls in building acoustics.