Durand D'souza

Data Scientist & Journalist

A graduate with a Masters in Physics with Astrophysics from the University of Leeds. Passionate about using data and machine learning to improve society. A quick learner with experience across the data science pipeline.
Portrait of Durand D'souza
durand1@gmail.com London, UK [Phone number in PDF] dldx.org durand101 dldx

Expertise & Skills

Machine Learning Natural Language Processing Data Visualisation Exploratory Data Analysis Strong numerical skills Web development Scientific analysis Critical thinking Fast learner

Programming proficiency

Python12 years
Javascript3 years
R3 years
D32 years
Fortran2 years
C/C++1 year
SQL1 year
Julia< 1 year

Work Experience

Data journalist & visualisation designer

Freelance

01/2017 - present Nottingham, UK
I researched, designed, coded and wrote two interactive visual essays for The Pudding publication. These essays may be found under the titles: "What 1.2 million parliamentary speeches can teach us about gender representation." and "We mapped out the road to gender parity in the House of Representatives".
  • Worked with the 50:50 Parliament charity and The Pudding to make the articles more compelling and thought-provoking while remaining politically neutral.
  • Taught myself complex machine learning pipelines and interactive data visualisation in javascript.
  • Wrote web scrapers to collect data from various sources and used a number of techniques to clean and organise all the data.
  • Used machine learning techniques such as LDA, POS tagging and T-SNE to identify broad topics within millions of parliamentary and congressional speech transcripts.
  • Optimised pipeline for limited memory requirements.
  • Created striking, custom visualisations to elicit readers into exploring the data themselves.
  • Wrote a compelling story to tie together speech analyses and interactive visualisations.
  • Designed a complete CSS style and article format from scratch with iterative design and user testing to perfect it.
Also developed several private visualisations for campaigners and have a number of articles in preparation.

PhD Student

Max Planck Insitute for Astrophysics

08/2013 - 04/2016 Munich, Germany
My thesis work at the prestigious Max Planck Institute focused on modelling the evolution of stars much bigger than our Sun using theoretical astrophysics computational simulations. The eventual goal was to use my results to infer the interiors of stars in our galaxy using lightcurve observations from space telescopes.
  • Improved decades-old FORTRAN computational code which solves first and second order differential equations.
  • Created new custom visualisations using python to analyse computed models in detail
  • Created new tools for myself and collaborators to manage simulation data and analyse it in detail in a modern environment.
  • Aided colleagues working on related projects, especially with technical problems.
  • Studied German to B1 proficiency.
  • Gained certification in C/C++ programming.
While I eventually decided not to pursue a doctoral degree due to various factors, my time at MPA taught me a lot about perseverance in high-stress environments and about how to communicate complex technical problems in a succinct and productive manner.

Education

Integrated Masters in Physics with Astrophysics (MPhys) - 1st class

University of Leeds (2009-2013)

Economic Modelling & Environmental Economics

Self-taught (2016)

A Levels - 3 As (Further Maths, Maths, Physics), B, 2 Cs

The Becket School (2007-2009)

Interests

Tackling climate change and inequality are two of the main challenges that drive me. I enjoy educating people about these issues using hard data presented in a clear & concise manner. I also volunteered at Oxfam for several years and have helped out in campaigns against climate change.

In my free time, I enjoy hiking, climbing, photography, and hosting friends for dinner.

Download as PDF