If you’re active on social media, you’ve likely seen (or even used) PopSugar’s Twinning app, which uses facial recognition to show what celebrity you look most like. This app, powered by Clarifai’s AI has gone viral thanks to social media word of mouth and a shout out from Ellen on The Ellen Show.
I used the #twinning app, and I have to say I’m flattered. pic.twitter.com/ezmp70xAXD
— Ellen DeGeneres (@TheEllenShow) January 7, 2019
But what happens when your app goes viral? Well, you better hope that the tech you built or the tools you invested in are able to scale with you. Luckily, as usage spiked last night, Clarifai was able to provide the needed support for the PopSugar team.
“We evaluated multiple facial recognition services and found that only the Clarifai API gave us the flexibility and control needed to implement POPSUGAR Twinning the way we envisioned," says Brian Sugar, POPSUGAR founder and CEO and co-developer of POPSUGAR's popular Twinning app. "We’ve also been impressed by how well their service scaled as the Twinning app went viral.
I want it that way. https://t.co/ZIWydihATn
— Lin-Manuel Miranda (@Lin_Manuel) January 7, 2019
Interested in building your own AI app?
Sign up for a free Clarifai account here to get started!
We’ve discussed previously the pros and cons of choosing to build an AI solution in house or buying an AI solution. As Clarifai’s own Mike Vaccarino wrote in a recent blog post on the topic:
There’s no need to worry about the cost of monitoring and scaling an IT backend infrastructure as volumes grow and technologies evolve that require more compute power. There’s also the additional benefit of being tapped into a network of data that far exceeds your own, and being able to apply the aggregate learnings from this data to your own unique use case.
This meant that PopSugar was able to launch their app without the hundreds of hours it would take to train and test a model on every western celebrity’s face AND had access to Clarifai’s infrastructure, enabling them to scale as usage went viral.
