When you have a team this awesome, you can’t keep an entire blog to yourself. That’s why we’re starting choozlechat, a series of blog posts with contributions from Choozle thought-leaders across every team. Have an idea for a choozlechat? Email us!
Megan Sullivan-Jenks, Director of Marketing at Choozle: We sometimes start these off with the most important questions. So, if you could eat one type of candy for the rest of your life, what would it be?
Claire Nessler, Software Engineer at Choozle: I would have to say Snickers. One, because I get hangry and “You’re not you when your hungry”, and two, I love it! Honestly, I would be content eating it every day.
MSJ: It’s true, you’re not you when you’re hungry! Now, onto the technical stuff — How would you describe Elasticsearch to someone who is unfamiliar with development and/or coding?
CN: Elasticsearch is a tool often used to help in areas where there is a high amount of data querying. It is powered by an information retrieval library which is highly optimized for speed. When a user searches for third-party data segments in the Universal Data Catalog, their criteria is ranked based off of complex algorithms to give them the most accurate results.
MSJ: Is Elasticsearch used in Choozle code? For what?
CN: Elasticsearch was implemented for the new Universal Data Catalog. When a user searches for data segments to add to their custom audiences, that search is now powered by Elasticsearch. When building this new catalog, we saw a great opportunity to speed up a slow query in order to have a better user experience. Elasticsearch is also used behind the scenes to store data. This is typically referred to as a log, which simply stores pieces of information. The information that is stored may consist of errors that have occurred when an ad group is saved to the database, or when a report is pulled.
MSJ: What benefits are Choozle users seeing inside of the platform due to Elasticsearch?
CN: For one, Elasticsearch makes searching simple. It removes the need to provide education around using quotations or parentheses when searching through third-party data segments. Elasticsearch handles the results based on the search criteria inputs, which allows for super-speed when searching and including or excluding data segments from custom audiences, and increased accuracy within search results.
MSJ: What are the pros of using Elasticsearch compared to the alternative(s) you and your team were considering?
CN: Truthfully, our team wasn’t considering any other alternatives when building the new catalog. Elasticsearch could be implemented quickly with a ton of features ready to go, straight out of the box. This solution handles searching and logging really well, which was our primary focus. Elasticsearch has been on my team’s radar for a while, and revamping the Universal Data Catalog was a perfect opportunity for implementation.