From Hackathon to Startup: The Creation of Rivrb Inc.
If someone had told me earlier this year that I would be starting a company based off of a hackathon project, I would not have been surprised; not because I am good at starting companies or anything like that, but, rather, because I am good at committing wholeheartedly to any endeavor I engage in. When the calls for racial justice reached a feverish pitch this past June, I did not know how to contribute towards the effort of creating a more equal and just world. As a Software Engineer by trade, I tend to overanalyze every situation in hopes of finding a best-case scenario, but in this case I did not know which path to choose. On the one hand, I could raise awareness of racial issues like my peers on social media, or, on the other hand, I could attempt to affect change in my own unique way. The problem I saw with the first option was that all the bases seemed to have been covered. What do I mean by this? Well, it seemed like everyone I was following was posting about the same issues. This led me to wonder: What new audience was I going to reach by posting the same content on my social media feeds? Yes, I recognize that the act of posting something on one’s social media feed shows that one cares about the issue enough to let their online community know about it, but some people are better suited at that than others. As a result, this led me to consider the second option. That’s when I found out about “NYC Coders Hack For #BlackLivesMatter” hackathon.
The hackathon was honestly a godsend, as I was not getting any closer to finding a way to contribute my talents towards the cause. NYC Coders is a meetup group geared towards creating a collaborative community for professionals looking to enter or who currently work in the tech industry. As their meetups had become virtual due to Covid-19, they decided to use their platform to host a hackathon geared towards creating software applications that could assist in the fight for racial equity. When I found out about this opportunity, I realized that most of my social media networks were not privy to this information, so I shared it on my Instagram account to raise awareness of the opportunity — thus, incorporating option one into my own unique plan.
At the start of the hackathon, there was an ideation period for participants to propose project ideas. It was at this point that I suggested the idea of creating a social media application that would aggregate users’ social media accounts and generate a feed that shows opposing content in hopes of eliminating online echo-chambers. The inspiration for this idea stemmed from my experience of the 2016 presidential election. During that particular election, everything on my social media feeds indicated that Hillary Clinton was going to win by a landslide, and when that didn’t happen I was truly shocked. It made me realize how powerful social media algorithms are at funneling people into echo-chambers. To my delight, there were other Junior Developers who were interested in working on an MVP for this idea, primarily Arianna Choza, Donovan Richardson, Reese Bowman, and our tech-shadow, Sanket Patnaik. Together with our hackathon mentors Omar Jameer, who functioned as our PM, and Marisa Ruiz Asari, who functioned as our designer, we were able to develop an MVP called De-Echo. It essentially asked a user for their political leaning, and then displayed the feed of a random Twitter account with an opposing political lean.
Some might wonder how did De-Echo relate to the Black Lives Matter movement. Well, at the core of the idea is the desire to get people of all backgrounds to understand one another’s perspectives. I believe that if people were to take the time to engage with the information that contributes to the fabric of other people’s worldview, then it might become easier to begin the difficult conversations that need to be had. In this vein, I thought our hackathon MVP was a decent first-step at implementing this idea. But I didn’t want it to end there.
When I decided to participate in the hackathon, I did not want to spend my time working on something that would just get ignored and thrown by the wayside within a few days or months. I wanted to contribute to something that would be useful to people years down the road. So after the hackathon, I proposed the idea of turning our hackathon project into a startup company called Reverb — as in the reverberations and echoes created by one’s social media imprint. Upon my proposal, only one member of the hackathon team, Reese Bowman, was interested in creating a startup centered around the hackathon MVP with me. This was on June 17th. It was go time!
I became CEO, and Reese became President. We needed a CTO, COO, CMO, and CDO — Chief Design Officer. I reached out to many of my closest friends and acquaintances whom I knew would be passionate about the idea, and who had the superb talent to bring this idea to fruition. At the same time, I started reaching out to industry leaders in tech, finance, entrepreneurship, and academia to gauge their interest in being advisers for this novel startup. From June 17th until August 1st, I had individual conversations with over thirty different individuals about the idea, and pitched it to over 70+ others via various virtual startup conferences I attended. The idea was picked apart from beginning to end, from business model to implementation, and by the time August rolled around the idea had changed dramatically.
First of all, I had changed the name from Reverb to Rivrb due to the unavailability of the Reverb.com domain and the availability of the Rivrb.com one; We had formed a legal business entity called Rivrb L.L.C. in the state of North Carolina, and we had found a COO, Seyi Olojo, CMO, Mackenzi Turgeon, and CDO, Von Storm. That only left CTO, a position that is really difficult to fill, so I donned the mantle in tandem with my other responsibilities. So back to the company idea. Through many conversations with our advisers and others, our team was able to realize that it’s very difficult to implement a business model based around showing people information they do not necessarily want to engage with. Moreover, we had decided to make our feed generating algorithm open-sourced, allowing interested parties to freely audit the algorithm and potentially contribute their own ideas and suggestions to the project. Furthermore, we didn’t want the algorithm to be dictated by the need to show Ads to users, a distinction that is necessary in order to ensure we don’t end up recreating the same echo-chambers we set out to dismantle, as Ads targeting requires an intimate knowledge of what people “like.” How do you find out what people like? By reinforcing the rate at which content they engage with shows up on their social media feeds. As we explored the idea even more, it became clear that the big problem we were trying to solve was content validation at scale.
What is content validation at scale? There are many entities out there working on content validation. The simplest way to explain it is the attempts of sites like WT.Social and FactCheck.org in validating news, but their implementation isn’t really at scale, as the majority of the validation is typically done by human labor. In fact, the clearest example of validation are verified, blue-badge accounts on Instagram, Twitter, FB, etc. that indicate that the individuals with said accounts are who they say they are. Now expand this to verifying news articles, then to verifying any article. How about verifying any social media content? It’s an immensely large-scale problem that is only getting more difficult by the day. I recently took a break from social media, and when I came back I was amazed at how intelligent Instagram and Snapchat filters had become. This is all fun and games, but what happens when deep fakes become intelligent enough to fake the voice, appearance, or writing of political leaders or individuals with authority? Moreover, I hope it doesn’t take long to imagine what could happen when unverified and false information is propagated in online communities fueled by algorithms seeking to target Ads at unsuspecting users?
To bring this idea to fruition, one of our advisers, Hussein Rashid, a Columbia graduate and Harvard Ph.D who has taught at Columbia University, suggested that we target our MVP towards college students. If we could provide them a platform where they could enter a link to an article and then be provided with other articles on the same topic but which express opposing viewpoints, we could then expand on that idea as we work to scale the validation process. With this in mind we started developing our MVP, and implementing the other aspects of our business.
To say the least, the idea changed dramatically from what it had been on June 17th, to what it evolved into. Understandably, Reese had other competing priorities and decided that Rivrb would take too much of his time, and so he stepped down as President. Slightly before this, it had come to my attention that high-growth tech startups should be incorporated as a Delaware C Corporation, and not an L.L.C. So with this new knowledge, Rivrb L.L.C. was dissolved and Rivrb Inc. was formed with four members on the Board of Directors.
It’s been an intense journey so far, with many ups and downs, but we’ve made some amazing progress in the almost six months since the initial conception of Rivrb. I’ve watched all the executives grow and shine in their respective roles. I’m truly excited to be working with the executive team we have, and leaning on the wisdom and advice of our advisers. There are many more initiatives that Rivrb is embarking on, like creating financially sponsored Rivrb clubs on college campuses — an initiative that has already begun at Columbia University with the recent selection of Jordan Ordonez-Chaguay as the first Rivrb Club President — and raising awareness of our open-source machine learning initiative. This is a four year long process that I’m excited to be embarking on, and which I wish you all to be witnesses and participants of. Keep up with Rivrb Inc. at rivrb.com.
Calvin Goah is the CEO & CTO of Rivrb, an early-stage tech startup that aims to be a leader in online information verification and community information literacy.