Curology’s March 2021 Hackathon

Patrick Minton
Curology Tech
Published in
4 min readJul 20, 2022

The tech team at Curology recently held a week-long, team-wide hackathon, an event that we’d like to repeat at least twice a year. Going into the event, we had three primary goals: to improve team building (especially in cross-functional ways), let people pursue their passion, and most importantly, have fun! To meet these, we decided on a framework with very little structure; we basically made a giant Excel sheet with project ideas and asked people to use it to sign up.

In the kickoff meeting, we had folks pitch their ideas to the larger group. These projects ranged from fun visual ideas (like creating a video that illustrates our first 100,000 shipments), process improvements (like writing some code to make it easier to prototype an API in TypeScript), metrics (automating the measurement of our website’s page speed), and even climate activism (what’s the environmental impact of the products we ship?).

Following the pitches, folks signed up to work on the projects, and we split up into teams. Participants weren’t just software engineers; we also had a bevy of data scientists, data engineers, product analysts, and site reliability engineers, so the projects covered a broad set of domains. The teams also had one other important task: to pick a team name, using the theme of trucking or truckers. Puns were heavily encouraged.

There were a lot of great projects! We can’t go into detail for all of them because of space reasons, but here’s a sample of the projects we worked on.

First 100k shipments video

Team Eastbound and Production’s Down (Eric Wong, Thonda Taylor II)

This team’s goal was to visualize the early history of Curology by tracking its first 100,000 shipments in a timeline video. It took Curology about a year to reach 10,000 shipments, but they scaled up to 100,000 pretty quickly after that. To create the video, Eric compiled the zip codes and the shipment arrival timestamps from our database, and fed the data into a 3D visualization library for maps and routes. We might further expand this into a dashboard that we can use internally to see where all of our shipments are at any given point in time.

Measuring environmental impact

Team Garbage Wreckers (Eric Haavind-Berman, Eric Zhou, Mike Li, Kaiyi Li)

It may have taken Curology a couple of years to reach 100,000 shipments, but nowadays we ship tens of thousands of boxes and bottles every day. This team was curious about what the environmental impact of all that shipping is, in both materials and carbon emissions. They set out to calculate and visualize how many empty bottles are in circulation (by weight and volume). Using that information, they were able to visualize how far shipments travel and estimate the CO2 equivalent of our impact.

The team then started to calculate the cost of potential solutions, such as reclaiming used bottles, or sending refills instead of full new bottles.

Customer sentiment analysis

Team “Cyber Truckers” (Zachary Caceres, Dominique Gordon, Thonda Taylor II, and Amin Vakhshouri)

This project used Machine Learning to analyze sentiment analysis on various social media sites. In simpler words, they wanted to ask “how are people perceiving Curology on Twitter and/or Instagram?”

They fed data from Twitter and Instagram into a pre-trained sentiment model on Amazon Comprehend, dumped the results to a Firebase Firestore, and then used the data to power a dashboard that would provide the end user a snapshot of how folks were talking about Curology on social media.

Takeaways

Overall, the team really enjoyed working on unique passion projects during our hackathon, and folks got to collaborate with people from other teams who they normally wouldn’t interact with quite so closely. A lot of very cool projects emerged, some of which might make their way into production soon!

Appendix A: Complete list of projects

Cyber Truckers: Customer Sentiment Analysis Dashboard (Twitter and maybe Instagram)

Eastbound and Production’s Down: First 10,000 Shipments Visualization

Velvet Thunder: Build solution for parsing SQL comments and structure from code

Dispatcher Patcher: One integrated Cli tool for DE quick admin tasks

Haulin’ A$$: Measure Core Web Vitals

Mother Truckers: Git: Renaming “master” branch to “main” in our repos

In the Merging Lane: TypeScript schema parsing and code generation

Garbage Wreckers: Measure Our Environmental Impact

Road Ragers Anonymous: Image comparison slider UI for dashboard

Lone Wolf Trucking: Model for calculating daily conversion goals to track pacing throughout the month

Automatic Transmission: Automate take-home interview grading

Monster Truck Rally: Improve the deployment queue in #hax

Wicked Wanderer: Add a search feature to the guides page

Curasourus Rex: Create a “Your Skincare Journey” Game

Fig Rig: Figma plugin to batch export content

Big Reg(ex): De-Identify and do Topic Modeling on Helpscout Messages

Originally published on April 22nd, 2021 at https://curology.com.

--

--