In Fall 2019, we released the first version of Localify as a locally-focused artist and music event recommendation app. A couple of months later, in March of 2020, COVID-19 effectively ended live music just as we were starting to gain a bit of momentum. We soldiered on and tried a few pivots but Localify got a bit stale over time and all of the students who had worked on it eventually graduated leaving behind a fair amount of tech debt.

In spring 2022, a very talented student name April Trainor suggested that we build a new backend architecture for Localify. This involved:

• switching our web framework (from Flask/Python to Spring MVC/Java) to be more robust and scalable
• designing subsystems (e.g., web scrapers, recommender system) as independent micro-services that communicate with internal message passing (RabbitMQ)
• developing better MLOps tools to help update and evaluate the performance of various recommender system algorithms

Another talented student, Griffin Homan, was tasked with updating the old React-based frontend to match this new backend. However, Griffin inherited code that had become somewhat bloated over the years. By the Fall, he decided that also wanted to rebuild the frontend from scratch to clean things up. At the same time, Localify.org was being used as a project focus in a marketing course at Ithaca College. This provided us with valuable user feedback and novel design ideas. Over the winter break, Griffin started working with Liz Richards on a Figma UI/UX prototype so that we could flush out many of these design ideas:

• starting with mobile-first design and implemented using React-Native
• improving the onboarding, especially for guest users
• giving artists and super-fans tools help us crowd-source artist and event information
• redesigning pages for artists, venues, cities, and events

In order to monitor Localify, we also thought it was important to build out an administrative dashboard. This would allow us to be more responsive to users, evaluate crowdsourced contributions, monitor our web scrapers, and measure the performance of our recommender system over time. The dashboard was initially designed by Griffin but has been largely implemented by Paul Gagliano.

Given that we blew up everything else, we decided to also rebuild our web scrapers (e.g., Spotify, Last.fm, BandsInTown, RateYourMusic) from scratch. Kieran Bentley was put in charge of this so that by the end the scrapers will be much more efficient, fault-tolerant, and work with our micro-service architecture.

Finally, Vicky Conrad was asked to create this very landing page and blog so that we could share updates, research, and other information about Localify.org.

Please stay tuned as we get ready to launch Localify 2.0 this coming July!

ps – And just for a historical record, here is a screencast demo of Localify 1.0:

Written by Doug Turnbull – June 7th, 2023