The Google Cloud Platform team recently hosted a live webinar and Q&A Session with LiftIgniter founder Indraneel Mukherjee to discuss the benefits of building a startup with GCP. During the webinar, he discussed what LiftIgniter does, gave a view into the GCP-powered architecture behind LiftIgniter, and shared LiftIgniter’s experience with GCP.
LiftIgniter, is a machine-learning personalization company. LiftIgniter harnesses a similar technology and approach as YouTube does with its video recommendation platform, which serves up specific new content recommendations based on the user’s interests in that moment. YouTube is serving up personalized content from a pool of billions of videos for the each of billions of visitors exploring its site at any point in time, and in order to achieve this personalization at scale, it leverages a powerful machine learning technology behind the scenes. Think of the hours we can spend engaging with video that is in alignment with our specific interest in that moment, it’s addictive. For media brands keeping visitors on their sites consuming more content is a critical component of a satisfying customer experience. For eCommerce and CPG companies, serving up merchandise that appeals to their consumers and delivering the right purchase is key to keeping customers happy. LiftIgniter enables companies to recreate that habit forming engagement with their content, articles and merchandise.
What’s different today is that massive breakthroughs in machine learning, along with advances in cloud computing solutions such as GCP and infrastructure improvements have made it possible for LiftIgniter to take its technology to market. In order for LiftIgniter to be able to deliver real-time recommendations on par with YouTube, it has to do three things at once: sophisticated modeling on billions of rows of data, create those models at scale, and do it all really fast, in less than a second.
Indraneel also touched on the fact that while machine-learning can do wonders, specifically when it comes to crunching large amounts of data, it can’t solve for everything. AI does not have the cognitive function and intuition of a human. For this reason LiftIgniter enables a dynamic partnership between machine-learning and human intelligence to create a real-time, human driven experience that machine-learning by itself simply can not replicate.
Indraneel highlighted LiftIgniter’s architecture and transactional scale to show its amplitude and power. LiftIgniter recommends tens of thousands of items every second, and processes over 20 billion events per month, much of this data being processed through GCP.
LiftIgniter’s decision to transition to GCP was an easy one, having previously worked at Google, Indraneel was very familiar with Google’s technology, especially their bleeding-edge machine-learning tools, that along with their demonstrated commitment to the cloud business led to LiftIgniter make the switch to Google Cloud. GCP’s support and commitment to helping startups through marketing exposure like LiftIgniter has kept us happy and why we plan to stay.
Big thank you to Google Cloud Platform for inviting LiftIgniter to partake in their webinar about building a startup using GCP.
