Engineering Spotlight: Applecart
Applecart is a company that deploys proprietary technology to run smarter advertising campaigns. Its Social Graph platform builds dense, accurate, and actionable maps of real-world relationships between individuals to leverage them for maximum influence and advertising impact.
We connected with William Fry, Applecart's VP of Engineering, to learn more about the engineering organization at Applecart and much, much more.
Applecart is hiring! Check out their BIZZpage for all of the job openings at Applecart.
Can you share a summary on what Applecart does?
Our core technology is a national-US social graph that leverages publicly-available data to map real-world relationships between individuals at scale. We use this graph in our two product lines: an enterprise offering that enables our clients to run smarter marketing campaigns inclusive of political campaigns and corporate advocacy and a direct-to-consumer app that compensates consumers for referring their friends to the brands they genuinely love.
What are some of the different technologies that the engineering team gets to work with and at what scale?
It’s simplest to break down the different technologies by use case. Right now we have three general activities on the engineering side: constructing the social graph and deploying it in our two product lines.
For the graph construction and large-scale advertising campaigns, we use Spark a ton given the sheer amount of data we need to process. To enable ease of collaboration and prototyping, Databricks is our favorite tool to use on this front. In addition, we use common machine learning tools as well as a few third-party vendors such as DataRobot for our predictions.
On our direct-to-consumer product line, we need to provide real-time services to our users, so we use a standard mobile stack. We have backend services written in Flask and Django which are deployed as Docker containers to AWS. The mobile app is written in React Native, enabling us to move quickly with apps on both stores.
What are some of the interesting projects that the engineering team is tackling?
We have no shortage of interesting projects. A few common themes that we’re tackling as of late include entity resolution, microservices, and moving to a real-time prediction paradigm with the D2C mobile app.
Entity resolution is important because we have dozens of different datasets and need to understand when two entities are the same despite differences in their features or whether they are truly different.
Our direct-to-consumer application is built with microservices, enabling our product teams to iterate quickly and independently. The common challenge with microservices is ensuring that contracts between services are clear and that we have strong visibility into performance.
Lastly, our newest product line provides real-time recommendations which is quite different from our historical work in a batch paradigm. This means we need to be able to quickly recommend brands or friends but still retain the ability to continuously improve the model. This has been a big focus of the recommendation team.
Does your engineering team have a chance to work on projects outside of their day-to-day responsibilities? For example - skunk work projects, open-source projects?
Yes! We take pride in in-office hackathons and time for professional development. The other day, we had a full day in-office workshop with AWS, a close partner of ours, to get team members trained up on deployment options for containers. More generally, learning, development and intellectual curiosity is what drives us. To support this, we allocate budget for team members to attend conferences and meetups as well as hackathons once every couple of months. I was very impressed by the scope of projects as the hackathon.
On the open-source side, we obviously use a lot of open source technologies so we encourage team members to contribute back. Whether it’s a streaming technology like Apache Pulsar or a React Native library, we take open-source seriously.
What is the culture like at Applecart for the engineering team?
When thinking of our team, I think they can best be described as driven, curious, and trusted. We value autonomy and ownership a lot. Given the pace at which we move and the novelty of some of our projects, our engineers are comfortable with ambiguity and all have a figure-it-out, can-do attitude. We trust each other to get the job done, but also prioritize taking the time to teach and learn from one another. We also take bagel Friday’s very seriously and do monthly team outings which can be anything the team wants to do like VR World or an arcade bar.
What can a potential employee expect during the interview process?
Our interview process is pretty painless. Hannah, our VP of People, has put a lot of thought into making this as seamless and feedback-oriented as possible, so she deserves all the credit here. For engineering positions, you will speak with a member of our recruiting team first. We want to ensure you have the opportunity to learn more about Applecart and we get to learn about your background, what you like to build and what’s most important to you in your next role. After this we have a short remote coding challenge. The goal here is to learn how you think and to ensure a strong technical baseline before proceeding. Following the coding challenge, you would come on-site and meet the team, see the office, and visualize what it’s like to be part of our team. The on-site includes a whiteboarding session and then a few sessions with Tech Leads and Product Managers because we think it’s important that you meet your immediate team and the greater team you’ll collaborate with.
Are you involved in any local tech organizations or Meetups?
Applecart is not currently formally involved with any meetups but a number of us are regular attendees of NY meetups. Our favorites include Data-Driven and Brooklyn JS.
Rapid Fire Q&A
What’s on tap?
Every Friday we have an in-office happy hour. Normally includes beer and wine with a healthy dose of Super Smash Bros.
Star Wars or Star Trek?
Controversial, Star Wars on the engineering team but Star Trek may take the prize across the entire organization.
iPhone or Android?
Android by a small margin
Coffee - hot or iced?
Hot, always brewing in the kitchen
Favorite employee perk?
In addition to your traditional desk set up with dual monitors, laptop and any accessories you need, we provide all employees with a several hundred dollar desk budget to customize their space and make it their own.
What TV show describes the engineering team’s culture?
Rick and Morty
What music is playing in your office?
It varies based on who’s DJ’ing. Both of our co-founders used to DJ back in middle school so there’s some stiff competition.
View from your office
Cleanest desk / Messiest desk (photos)
Rekha is awesome. She joined the team just three months ago and has really hit the ground running. She’s a unicorn of an engineer, bringing the data engineering skills needed to productionize models with the data science skills for model selection and evaluation. At Applecart, she sits on the recommendation team for our direct-to-consumer app. She runs point on the different models we use to recommend brands and friends as well as the integration of those models with our prediction and retraining framework.
Weston is everyone’s favorite teacher at Applecart. He’s incredibly knowledgeable and loves learning about new technologies, frameworks and concepts and sharing them with the team. At Applecart, he’s the Tech Lead for our direct-to-consumer app. As Tech Lead, he helps bring all of the backend services to life and is an expert in all things Python, Django and databases. He’s personally contributed to our payments and validation modules, while helping design and mentor those implementing much of the other components of our stack.