Please share the core guiding principles that you have learned about AI after having worked at some of the top technology companies in the world?
AI essentially is computer programming that learns and adapts from user behavior. The potential waiting to be unleashed from AI to transform the world, as we know it, is immense. At Hike, we are deploying AI to redefine social connections and empower consumers to supercharge expression. We are leveraging ML libraries like TensorFlow and some amazing homegrown talent at Hike, but some core principles that continue to guide me are:
– Focus on high impact AI projects, that can deliver solid business ROI to the company.
– Delivering through agile methodology; deliver initial output and get buy-in from key stakeholders to stay invested.
– Data collection and cleaning critical to obtaining high-quality results.
– Cross collaboration with business, product, tech essential for AI & ML teams to deliver meaningful impact to the company. Hence, this collaboration should be at the core for ensuring success of AI projects.
– Output can be probabilistic. Hence, that needs to be taken into account while deploying for real-world applications.
– Output from the AI model needs to comprehensible to give rational explanations to the domain experts.
What kind of AI roadmap have you set for AI in Hike?
At Hike, the belief is that technology should wrap itself around people and not the other way around and that AI is definitely leading the way to actualize this vision. We have achieved great strides at pushing ML models on the device to help users find stickers they like in real-time and we are on track to enhance the hyper-personal and hyper-local user experience even further. AI & ML is at the core of the company. We’ve used it in amazing ways to scale expression like never before. Our areas of research and application include Natural Language Processing (NLP), Computer Vision and Social Network Analysis (SNA). From creating the largest library of conversational stickers of over 1 million stickers to pioneering research and development in the field, we’re excited to enable the next leap in the ecosystem.
We are pushing the boundaries and building a new social future with AI at the core of it. The world is evolving at such a rapid pace, it’s only natural for social mediums to do so as well. Social connection is a core human need, important enough to dedicate a mission to. We believe the timing couldn’t be better. We’re excited to do this with homegrown AI & ML talent.
Benefits from using AI at Hike?
AI has opened the door for tremendous innovation, to further improve things that we today assume have reached their limits. So much more is possible today that wasn’t even just a few years ago.
From building radically unique products at the intersection of Product, Design, Engineering and Art to cutting-edge work across NLP, Computer Vision and more. We have been able to make some significant leaps here— from showcasing at renowned global platforms such as TensorFlow World, IJCAI & ECIR to live application of research through our powerful sticker recommendation and discovery features. It’s also interesting to note that Hike is the only player that is using NLP to solve for local languages at a mass scale.
With a huge focus on research-led innovation, we have also led partnerships with local academia such as Indraprastha Institute of Technology Delhi (IIIT-D). Our unique advantage of enabling locally relevant research, especially for a diverse market like India, has been key for these partnerships. Cultivating a culture of AI innovation has immense benefits both for Hike as well as the ecosystem. As an initiative in this direction, we have recently launched the Hike Patent Programme, which not only incentivizes Hike employees with rewards and grants but also lends legal and market guidance to prospective patent filers.
What are your views on having a robust supportive IT infrastructure for AI, for example, the required amount of data integrity and consolidation/ data lake or warehouse, etc.
Data is the building block of Artificial Intelligence (AI). Thus, a robust IT infrastructure enabling a solid data storage strategy equipped for the burgeoning data growth is pivotal to AI’s success in any organization. Enterprise Data Warehouse and Data Lake are needed to support diverse data sets that can be used for AI applications. Data flows need to be robust and scalable with high performance. Efficient infrastructure including CPU and GPU clusters are needed to build scalable and complex AI applications for large scale data. Streaming data infrastructure such as Kafka pipelines are needed to provide AI solutions leveraging a combination of online and offline data. Last but by no means least, data security and privacy powered by a robust AI infrastructure are key to ensure the confidence of the user and customer enterprise.
Please tell us about Hike’s participation in the Tensor Flow World 2019 conference?
Hike’s vision is to build a new social future where social products are joyful. A future which will see products built around people and not the other way around. To enable this vision the company is guided by 6 intrinsic principles, which you can read about here. A key principle here is advancements in tech & AI. These advancements enable us to push richer and more personalized experiences in real-time to the users. This is where events like TensorFlow World 2019 play a key role. This is the first event this year that saw the best minds in ML come together and discuss the role and impact of TensorFlow & TensorFlow Lite in the community. It was great to represent Hike as the only ‘Made in India’ company and deliver a keynote on our AI-driven innovations & the unique aspects of how we deliver sticker recommendations. To further elaborate:
With TensorFlow and TensorFlow Lite, the kind of work Hike has accomplished is pretty unique in the country. We have brought in the hyper-personal and hyper-local elements by addressing 8-9 Indian languages and by doing sticker recommendations in a very effective fashion at a low latency of few milliseconds.