Hi there, I'm Paul Fletcher-Hill, a Computer Science and Economics student at Yale. I'm currently building PatientBank. In the past, I've worked at Goldman Sachs, Upstart, Artsy, and the JHU Applied Physics Lab. I'm passionate about health care tech, quantitative finance, and elegant design (of all varieties).

PatientBank

PatientBank is a platform for sharing medical records with hospitals, doctors, and family members.

Many hospitals have gone digital, implementing electronic medical records. But these systems do not communicate and are not interoperable. Doctors waste time and money gathering records for new patients. And individuals lack the ability to access and manage all their information in one place, especially when seeing multiple doctors.

PatientBank automates the process of requesting medical records from any doctor. We store health information securely online and make it easy to share records with family members and doctors. Because health records stored in PatientBank are controlled by the patient, hospitals don’t have to interface with other hospitals—they access all the information they need via PatientBank. The platform eliminates the need for (and cost of) interoperability.

I am a co-founder and CEO of PatientBank. If you're interested in learning more, please visit patientbank.us.

Work

Goldman Sachs

Equities tech summer analyst, Summer 2014

Built a tool for analyzing usage of Falcon, the primary equities and fixed income trading platform. The work included static code analysis using reflection in Java, a log parser, and Elasticsearch. The product is used by traders or technologists to easily analyze user activity and identify inefficient or unnecessary features of the trading platform.

Upstart Network

Software Engineering Intern, Summer 2013

Contributed to Upstart’s web application with a revamped user sign-up flow, improved upstart profiles and calculator to assist investors when composing their portfolios of upstarts. I also worked on Upstart’s pricing algorithm, estimating a user’s future income stream to provide backer’s with a price per 1% of income.

Artsy

Software Engineering Intern, Summer 2012

Co-built two applications: the Artsy internal admin site and the content management system for gallery partners. The admin site automated a number of marketing and promotional tasks for the Artsy sales team, and the CMS allowed galleries to upload artworks directly to the Artsy platform.

Johns Hopkins University Applied Physics Lab

Operations Intern, Summer 2010 - 2011

Analyzed Solid State Recorder data and performed operations for the Payload Operations Manager for the NASA MESSENGER Mission to Mercury. The mission launched in 2004 and the spacecraft entered orbit around Mercury in 2011, while I was at the lab.

A more lengthy description of my work experience can be found on my LinkedIn profile

Other stuff

At Yale, I spent almost three years working on TEDxYale. I helped organize the inaugural conference during my freshman year, led the organization as co-curator the next year, and advised the team throughout my junior year. My favorite talk from my years working on TEDxYale is Michael Frame's touching finale to our "Solve for y" 2013 conference.

I haven't done much blogging or writing, but I recently published An Open Letter to the People Who Brought Us HIPAA on the Health Care Blog. The piece discusses the HIPAA Privacy Rule, specifically how providers are given the right to charge individuals for access to their medical records.

As an independent study in the Fall of 2014, I worked on a project called the PatientBank Health Graph. Built from a few simple components, the Health Graph is able to integrate a patient’s medical information from across the health care system and represent it in a meaningful and scalable way.

For the final paper in CPSC 638 Database Architectures in the spring of 2015, I studied current approaches to large-scale graph processing. The paper, Large-Scale Graph Processing with MapReduce and Google Pregel, compares two paradigms in analyzing graph data: MapReduce and Bulk Synchronous Parallel (via Google Pregel).

In the spring of 2014, I wrote a paper titled Computing Shapley Values in the English Premier League. In 1953, Lloyd Shapley introduced the concept of a Shapley value for coalitional games. Using four convenient axioms, the innovation allows the calculation of a unique distribution of the game’s surplus among its players, representative of their marginal power. The Shapley value has successfully been used in many political and economic games. In soccer, like many other sports, rating players is often a shallow and arbitrary process. Simple metrics, such as goals, shots or saves, are easily digestible, but they hardly encapsulate the actual value of a player while he is on the field. The paper attempted to apply the Shapley value framework to soccer teams, specifically the 2013-2014 Arsenal Football Club in the English Premier League, with the hope that a more nuanced, teamwork-based measurement of a player’s contribution would arise.