Kia ora, I’m Ari

I joined Harmonic in 2016 as a Data Scientist where I was subsequently promoted to a Data Science Manager in late 2018. During my time I have been able to play a role in a number of projects, including large-scale shiny web apps for some of New Zealand’s most-visited websites, ethical machine learning projects to better improve business processes and research and innovation around bleeding edge data science technologies.

Take a look at the Harmonic website, for examples of the work I was apart of.

Before it was cool and fashionable for every news outlet to have a few maps & graphs at the the bottom of their COVID-19 related news articles, there was COVIZ: COVID-19 in NZ

Automatically scrapping the data from the Ministry of Health press releases, I mapped the early days of our COVID-19 journey, and presented the data in a way the hadn’t then been publically done

The application is currently still alive, however it hasn’t been updated for sometime.

The Engineering Revue is an awesome show put on by the University of Auckland Faculty of Engineering at SkyCity Theatre. Established in 2008, the Revue is an annual event drawing on the collective talent of dozens of students, from bright-eyed first years to world-weary post-graduates. Every aspect of the show, from writing and performance, through choreography and costume, to lighting and sound, is done by students.

I was involved as a co-director of the event for two years, in addition to another 2 years as a crew member.

More information about the show (which only gets better and better ever year) can be found at engrevue.co.nz

As part of a team of engineering students we designed and built a functioning wind turbine. Competing against other teams from our department, we tested our turbines in the University’s twisted flow wind tunnel.

We designed our tubines by building mathematical models of how turbines operate using fluid dynamics equations, building computational models, and writing code to iteratively improve the design of our turbine.

Using our optimised computational design, we physicaly made our turbines using laser cut perspex blades, and a machined central hub.

As part of my engineering thesis, I looked into how Markov Chain Monte Carlo (MCMC) simulations could be used to perform inverse paramater estimation of geothermal models of the Wotten-Waven geothermal reservoirs in Dominica.

The work involved configuring large amounts of computational resource to run geothermal models, and a serious understanding of both the statistic the underpins geothermal models, as well as how this interacts with the dynamical systems of a geothermal model.

In one of my first attempts at using an operations research approach to solving a problem, a small team of friends took on the challenge of creating an ‘optimal’ working schedule for Air Traffic Controllers at Auckland Airport.

After ensuring we understood the requirements for safe and effective staffing levels we took to work to look at how we could minimise the number of controllers we would need on duty at any given time, such that all staffing level requirements were met.

We looked at how our schedule would cope with air traffic in real time, under a variety observed days of actual air traffic. The results of our simulation can be seen below.