darkgreener.comAnna Powell-Smith

Hello! I’m a public-interest technologist. I am currently the director of the Centre for Public Data , a new non-partisan research and advocacy organisation. We work to strengthen data provisions in policy, and campaign to fill gaps in important public datasets. Before that, I was chief product officer and senior developer at Flourish , a successful British data visualisation startup. Before that, I was the founding technical director at the University of Oxford’s DataLab , which uses data to help doctors make better decisions. Write to me: [email protected] . Some past projects As a data scientist and developer, I worked on many technical projects using combinations of code, data, stats, maps and visualisation. By choice I worked in Python (pandas, Jupyter, Django), JavaScript (D3, Node), and Postgres/PostGIS. Mapping money laundering In 2015, I helped Private Eye map all the land in England owned by overseas companies, revealing London property being used for extensive money laundering via tax havens. The map was praised in Parliament and commended at the British Journalism Awards. Mapping land ownership From 2017 to 2019, I was the technical lead at Who Owns England , a project which investigated land ownership. I led the data and mapping work.Mapping house prices In 2017, I used new data on 6.2 million house sales to produce the first definitive data on prices per square metre of houses in England & Wales. The most expensive postcode in the country is SW1X, in Mayfair: check yours here . Doctors’ decisions While I was the tech lead at Oxford University’s DataLab , I was the sole developer on the beta of OpenPrescribing , allowing fast search and arbitrary visualisation of a billion rows of NHS prescribing data. Dress sizes I decided it was too hard to work out your dress size in different shops. So I built What Size Am I to make it easier. It was featured in the Guardian , Daily Mail , Economist , Wall Street Journal , BBC TV and many other places.Opening up Domesday Book Back in 2011, I put the Domesday Book (the original government dataset) online for free. Open Domesday maps every town recorded in the 11th-century Domesday Book, and has been visited by over one million people. Baby names I made a data toy to search English baby names, and wrote about the trends - Ava is up, Jordan down. My findings were featured in the Guardian , Telegraph , Economist , and on the Today programme. A few more: mapping UK farm payments , visualising the Premier League , calculating dress sizes (still my only project to make both the Wall Street Journal and Daily Mail ), a hedonic analysis of commuting times , and visualising statistical uncertainty . Football scores I wanted to understand football a bit better, so I visualised 20 years of the Premiership. It got a mention from the legendary Edward Tufte . House prices I wondered how travel time to London affected house prices. So I calculated it for UK stations and graphed the commuter zone visually. Search it to see your nearest station, or find out more about it from Nestoria . Shoe shopping I got frustrated by not being able to find the kind of shoes I wanted online. So I built a search engine for shoes. You can use it to search more than 20 UK stores, from Anthropologie to Zara. Taking notes Back when Android was brand new, I decided what it really needed was a notebook app with 18th-century fonts. More than 300,000 people downloaded the app, before I retired it in 2012. They gave it an average rating of 4.3 stars. Street problems I used to be a volunteer at democracy charity mySociety , and I wrote an Android app for the award-winning FixMyStreet . It made it easy to report problems in your street. An open-source project, it was used as the basis for versions around the world. Visualising uncertainty Funnel plots are a great way to show uncertainty visually. At an NHS Hack Day , I built a simple tool to turn spreadsheets into funnel plots. The code is MIT licensed and available on GitHub .Some past clients As a freelance developer from 2009-2015, I worked with large companies, non-profits, government and startups. Corporate clients included Marks & Spencer , RAND and Morgan Stanley . Startup clients included Axon , Nestoria , ElasticHosts (later acquired). Non-profit clients included Greenpeace and the Government Digital Service . Bio I have an MPhil in computer science from Cambridge, specialising in natural language processing. I worked in finance before becoming a developer. Find me on Twitter , LinkedIn , GitHub , or drop me a line: [email protected] . Speaking Past talks: include 11 Things You Didn’t Know About D3 at both OxfordJS and LondonJS in October 2013. Another talk about D3.js at GOTO Aarhus in October 2013. D3: Journey to the Source at the London D3 User Group in March 2013. D3: Data-Driven Delight at Front-End London in January 2013. Ten Things I’ve Learned about DataViz (pdf) at Hacks and Hackers London in July 2012. Opening up Domesday Book (pdf) at Wikimedia UK Annual Conference in May...