When we use Amazon Music and say, “Alexa, play some good music”, it feels almost magical that Alexa knows exactly what kind of music we would like to listen to. However, it’s not exactly magic.
What goes on behind the scenes is a medley of some of the most advanced AI-enabled recommendation systems. The AI cross-references your music taste with the best songs from a catalog of more than 90 million songs. The algorithms factor in a multitude of criteria including a deep understanding of how customers of different ages and backgrounds across different countries listen to music in different languages.
This requires cutting-edge technology to make it all happen in a few milliseconds on voice assistants. Voice recognition and assistance are among the most complex applications of data analysis and machine learning as we know them today.
Alexa, which is Amazon’s AI-powered voice assistant, is backed by a team of exceptional data and AI engineers. This team, led by Ashlesha Kadam, uses a combination of customer anecdotes, data, and voice technology to make high impact decisions including what and how Alexa should respond to different music-related requests.
She was the state board topper of Maharashtra in the year 2006. Subsequently, she secured her bachelors in Computer Science at BITS Pilani, went on to complete her MBA from IIM Bangalore, and started her post B-school career in the space of online sales and e-commerce, leading the online sales channel and operations for Allen Solly, a top premium apparel brands in India. Further on, she moved to Cisco Systems, the leading networking company in the world, in a core internet security team as a software developer.
Today, Ashlesha Kadam leads the global Amazon Music product team that builds music-listening experiences on Alexa and Amazon Music apps (web, iOS, Android) for millions of customers across 45+ countries.
She works on crafting the online music streaming experience ranging from very specific and niche requests (try “Alexa, play some French rap from the ’90s”) to even the vague ones (try “Alexa, play some workout music”) on Alexa. She has led several product launches for features that have made the music streaming experience on Alexa, unique and the most advanced among other voice assistants.
Many of her product launches have generated positive PR for Amazon Music in leading publications like VentureBeat, Android Central, VoiceBot.ai, Pocket-lint, Tech Advisor, and more.
Ashlesha’s area of expertise and passion is to use AI to build advanced musical experiences on voice assistants. She has an in-depth understanding of voice technology, including
Automatic Speech Recognition (ASR): alludes to translating the speech that a customer makes into text
Natural Language Processing (NLP): includes understanding the semantic meaning of a request (e.g. if someone says “I like this”, knowing what they mean by “this”)
Text to Speech (TTS): refers to the speech that a voice assistant makes (e.g. when Alexa says something like “Here’s a workout playlist you might like” before starting actual music) and most importantly,
Music recommendations: means using all the signals that customers provide in conjunction with what’s known about their past tastes and preferences and using that to provide the best, most relevant and contextual music recommendations for every customer in every moment.
Upon being asked about her stand-out qualities, Ashlesha says, “As a product leader and a former software developer, I have realized that I am uniquely equipped to understand the art of the possible on the tech side, ask the right questions to understand complexities, and create products that impact millions of users globally.”
Besides being passionate about using AI to improve customer experiences around the world, Ashlesha is an equally passionate advocate for women in tech. This year, she is serving as a co-chair for the Human Computer Interaction (HCI) track in 2023-24 for Grace Hopper Celebration (the biggest tech conference for women in tech with 30K+ participants across 115 countries).
With the recent advancements in Large Language Models (LLMs) that are making possible what was considered almost impossible only a few years ago, Ashlesha is excited about the new possibilities that LLMs unlock for voice assistants. She believes that with LLMs, voice as an interface can become an even more powerful, ubiquitous, yet natural interface for humans to interact with technology.
