Adelyn Flowers

Stack: Python, Spark, Airflow, AWS

As a software engineer embedded within a data science team I build apps, packages and CLI tools to help data scientists focus on work that matters. This includes a heavy dose of writing ETL pipelines and managing infrastructure like package servers, databases, and open source BI tools.

Notable Projects

• the miss report. A web app that identifies shortcomings in our video recommendations, including automatic outlier detection and daily emails with suggestions for A/B tests. Suggestions made by the miss report have improved customer product performance by over 30% this year.
• uvh. A desktop application that simplifies and aggregates information from multiple internal APIs in a clean and functional interfaces. This app saw universal adoption amongst customer-facing employees, and substantially reduced time to ticket resolution.
• emr-utils. A python package and CLI tool to simplify launching and interacting with Amazon EMR. Allows data scientists to define and save their most common configurations, easily utilize spark connect for local development, and version control EMR configurations for future use. This package has greatly simplified day-to-day work for the team.

Stack: C++, Python, Javascript

On a small team dedicated to preserving California citrus from a frightening disease, Huanglongbing. I took the lead on designing and coding a model of disease spread. The model allowed for the testing of multiple intervention strategies, cooperation levels, grower preferences, and economic variables. I used these results to lead author a research paper, whose results we used as the source of outreach to CA citrus growers. This work continues today, with the original model still serving as a cornerstone for new research.

Notable Projects

• hlb_model. A cellular automata model of disease spread. Written from scratch in C++, the model was built to be run in parallel on HPCs that the team received from a grant.
• hlb cost estimator. An interactive outreach website that helped farmers learn how to slow the spread of HLB and plot how different intervention strategies that would impact their short and long-term profits.
• scheduler. A python package/CLI tool for scheduling job submission to our IBM HPC. Abstracted technical details from team members with minimal programming experience.

Stack: C++, C#, Python

Over the three years I was an intern at Quandary, I got to get my hands dirty writing customer-facing code. Initially brought on to help write C++ programs for the microcontrollers that served as the core escape room experience, my responsibilities expanded to include writing employee facing software and automating arduous business tasks.

Notable Projects

• hint puncher. Named for the punchy sound that customers had come to associate with receiving hints, hint puncher was an all-inclusive piece of game management software written in C#. Hint puncher replaced a previous combination of 4 paid pieces of software, allowed employees to manage multiple games at once, and provided an exact view of what customers saw in the room for ultimate control. Hint puncher remained in use for 5 years, dramatically increasing employee productivity and customer satisfaction while decreasing onboarding time.
• the trunk. An all-in-one escape room that fit inside of an antique trunk and was used to promote the business at street fairs and other events. More than 8 microcontrollers were included throughout the trunk, powering everything from LED lightstrips, to motors, to pressure sensors and more. The trunk was wildly popular at every event it appeared at.