Dive
Dive was a multi-domain recommendation engine that helped users easily find derivative content for their favorite songs, movies, books, etc., and better connect with their friends over shared tastes.
No matter the task, you learn most by actually doing rather than simply watching or studying without application. Dive was the embodiment of this. In 2020, my co-founder & I had an idea that we wanted to bring to life. After dancing around it for some time, we thought, "Why Not Us?" and built a company that would change both of our lives.
"I hear and I forget. I see and I remember. I do and I understand."
-Confucius
Backstory
I've always had a deep love for music. As a child, I played four instruments and was in a Jazz band. Nowadays, my most important ritual is listening to a new album every day, and my favorite hobby is producing music with friends. In college, I met a friend who had a similar love for art, but for him, the medium was movies. He was a history buff and could probably name every post-1900 war movie or documentary.
While we bonded over our love of art, what was more important was our love of sharing the media that meant so much to us with our friends. With this in mind, we set out to create a tool that could facilitate this with ease, regardless of the media type. The three pillars of our project were the following:
Our love of media and sharing media with our networks
The college environment - College is a unique environment where media is routinely experienced with others via parties, movie nights, study sessions, etc.
An emerging prevalence of content-curation influencers - With increasing access to streaming content and the inability to parse through it all, we noticed a rise in social media influencers creating communities centered on content recommendations. Our tool would be perfect for facilitating this.
Project Overview
With a clear mission, our core goal was to build an app that would allow users to share six types of content (Songs, Movies, TV Shows, Books, Articles, Podcasts) regardless of the streaming platform they came from. This was supplemented by the additional goal of creating a community of media lovers eager to share their favorite art!
Unique Factors / Challenges
Across four years, our team encountered countless challenges. Down below, I've highlighted a few
Technical Expertise - Neither my co-founder nor I had prior experience coding. After outsourcing the initial development of our MVP, we learned to design and code, ultimately releasing a V1 built entirely in-house.
Inconsistent media metadata - One of the core foundational issues in relating content across media types is metadata inconsistency. For example, songs are often described by tempo and artist, whereas movies are often described by director, theme, etc. To overcome this, we developed a universal tag system that allows users to semantically describe the significance of their content and relate it to other types of content. This was the basis of our recommendation engine.
Handling of streaming platform hosting - With user posts in the form of streaming links, we built a parser that correctly ingests and displays media of any type via their streaming links. This was difficult as operations vary widely across the streaming landscape.
Fundraising - Without formal institutional backing for our first few years, we sustained the company by raising capital from angel investors & startup funds and participating in pitch competitions.
Achievements & Outcome
Project Achievements
Creation and release of the Dive mobile app on iOS
3K+ users across seven states and 10+ college campuses
Completion of a marketing campaign with Buzzfeed
Fundraising over $400K
Development of a Multi-Domain recommendation engine powered by a 16,000+ Node Graph database
Acceptance into the Microsoft for Startups program
Acceptance and participation in The Batchery Pre-Accelerator Program
Personal Achievements
Coding expertise - Learning how to code in Swift and Cypher
Entrepreneurial experience - Firsthand experience of taking not just a product but a business from 0 to 1
Team leadership - At our peak, our team consisted of 10 members, ranging from paid interns to full-time contributors
