Jacob Williams — Software Engineer

I helped maintain processing pipelines and search indexes for a web application that allows legal professionals to upload large data dumps and find which documents within them may be relevant to a particular legal matter.

On the search team, I rewrote our system for doing bulk updates and certain expensive queries, in order to eliminate an expensive-to-maintain custom Elasticsearch plugin. This also included performance testing and monitoring to ensure the new approach did not degrade user experience.

I resigned in 2023 to focus on academic pursuits.

I helped maintain and enhance Bloom, an ecommerce solution for wineries.

I designed and implemented a custom task scheduler for our background jobs, to give us greater flexibility in dividing processing resources among clients.

I worked on a variety of services/APIs, processing pipelines, libraries, and web apps, mostly related to products that help hospitals (and other healthcare providers) process and access patient and personnel data. I was involved in development, architecture design, and planning.

I saw a number of overlooked opportunities to improve developer productivity, and built tools that became popular internally. For example:

• One developer noted that a CLI tool I made to automate some common manual bureaucratic tasks reduced the time it took them “from ~45 minutes to ~6 minutes”.
• I created a browser-based tool called scrimp to fill the same role for Apache Thrift services that tools like Postman fill for REST services. This immediately became popular with both developers and solution designers within the organization.

I did substantial design and implementation work on a scalable search platform called Brahe, implemented as a Solr plugin that pulls data from HBase, along with a set of Java-based libraries and services, a Rails-based management app, and (years later) a Mesos framework to simplify operation of the platform.

I resigned in 2019 to enjoy some time off.