index.html

---
layout: default
title: Home
---

<h1>Devon Brackbill</h1>

<h3>About me</h3>

I earned my PhD from UPenn, where I was a member of the <a href="http://ndg.asc.upenn.edu/">Network Dynamics Group</a>. My academic research in Computational Social Science has been published in <a href="http://science.sciencemag.org/content/360/6393/1116"><i>Science</i></a>, <a href="http://www.pnas.org/content/early/2017/06/06/1615978114.abstract"><i>Proceedings of the National Academy of Sciences</i></a>, <a href="https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0237978"><i>PLOS ONE</i></a>, and <a href="/academic-research">elsewhere</a>. See <a href='academic-research'>academic research</a> for more details.

<h3>Current work</h3>
<ul>
  <li> <b>Data Science at Amazon Robotics.</b> I am currently a data scientist at <a href="https://www.linkedin.com/in/devonbrackbill">Amazon Robotics</a>. I'm working on improving robotic fulfillment and transportation buildings, all toward the goal of optimizing a logistics network that handles <a href="https://www.theverge.com/2018/1/2/16841786/amazon-prime-2017-users-ship-five-billion">billions of items each year</a>.</li>
  <li> <b>LP.</b> I am a limited partner in funds focused on software (typically SaaS) and artificial intelligence. For inquiries here, my email address is <b>< first-name >.< last-name >[at] gmail [dot] com.</b></li>
</ul>

<h3>Previous work</h3>
<ul>
  <li><b>Instagram content moderation.</b> As a data scientist at Instagram, I worked on content moderation, including the measurement and automated removal of objectionable content that violates <a href=https://www.facebook.com/communitystandards/>FB's community standards</a>. The problems in this space lie at the intersection of free speech and creating safe online communities.</li>
  <li><b>Health care.</b> Before that, I was a data scientist for a hospital working on population health. Our goal was to manage patient care to prevent them becoming sicker and incurring more expensive and intensive care. I built predictive models to forecast all-cause inpatient readmission risk and other events to reduce unnecessary patient care.

    build predictive models so that we could manage a panel of patients. Our goal was to manage patient

    worked at a hospital building predictive models for population health management.</li>


</ul>

<br><br>

<img src="/img/me2.jpg" class="img-thumbnail">

<a href="/academic-research"><h2>Research</h2></a>

My <a href="/academic-research">academic research</a> sits at the intersection of the social sciences, physics, and computer science. I am interested in how individual actions give rise to collective phenomena in large populations. Applications include the complex interplay of social forces that produce <a href="http://www.pnas.org/content/early/2017/06/06/1615978114.abstract"><b>collective intelligence</b></a>, <a href="https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0237978"><b>group problem-solving</b></a>, <a href="http://science.sciencemag.org/content/360/6393/1116"><b>consensus formation</b></a>, and <b>social norms</b> in human social systems.
<br><br>
To understand the dynamics of complex systems, I rely on agent-based models to gain theoretical traction on a social behavior. Then, I test these ideas using online experiments to see if these theories hold true.
<br><br>
See the <a href="/academic-research">Academic Research</a> section for more information.
<br><br>
<img src="/img/murmuration.jpg" class="img-thumbnail">

<a href="/blog"><h2>Blog</h2></a>

I have an occasional <a href="/blog">blog</a> where I write about research projects and interesting things like
<a href="/2017/06/12/collective-intelligence/">collective intelligence</a>,
<a href="/2020/09/01/causal-discovery/">causal discovery (or how you can learn the causal structure of a system)</a>, and hopefully more things!