Harsheel Singh Soin is a Senior Data Scientist in the Content Applied and Research Science team, helping to make content discovery across Audible surfaces ultra-personalized and enjoyable so that customers can find listens perfectly suited to their tastes. Find out more about the exciting work he and his team are doing as technology continues to evolve with AI.

What is life like at the hub you’re aligned to?

I’m at our Newark headquarters, where my colleagues and I participate in local volunteer activities, like meal preparation at local charities. We explore neighborhood restaurants for lunch and organize occasional team outings.

I've led technology workshops for children at Audible’s family events, making concepts like AI accessible through creative, hands-on activities. I also co-organize internal research symposiums that create platforms for knowledge sharing across teams and I mentor interns and new team members to help them navigate our organization. I've also taught data science fundamentals at a Newark school, connecting with local students interested in technology careers.

These experiences remind me that our technical skills can make a positive difference, whether through direct community service or by creating learning opportunities for others.

How does your job contribute to Audible’s mission to surprise and delight listeners around the world? What is your biggest source of inspiration?

My work helps customers get better recommendations so they can discover titles they might never have otherwise found. I help develop algorithms that analyze book content and predict which titles might resonate with different audiences, aiming to spotlight hidden gems alongside bestsellers. My biggest inspiration comes from seeing our technology connect listeners with stories that entertain, educate or inspire them in meaningful ways.

How does the work you do affect listeners or your colleagues across different marketplaces worldwide?

I've helped expand our recommendation systems and content evaluation frameworks to international marketplaces like the UK, Australia and Canada. I've worked on adapting discovery experiences for different marketplaces and supported the launch of customer feedback mechanisms in multiple countries.

Working with diverse content requires understanding behavioral and cultural nuances in how different marketplaces consume audio entertainment. I've learned that while the core technology remains similar, implementation needs flexibility to accommodate regional preferences and listening habits.

What was your first Audible listen? What inspired you to check it out?  

My first Audible listen was Sapiens by Yuval Noah Harari. It offers a fascinating perspective on human history, exploring how our ability to create and believe in shared narratives—including nations and economies—has shaped our species' dominance. Derek Perkins' narration brought clarity to the complex concepts and made my commute both enjoyable and intellectually stimulating.

Can you share a bit about your career journey at Audible?

I joined Audible more than six years ago as a Data Scientist II working on content evaluation models that predict book performance before release. Over time, I expanded my focus to include customer-facing personalization and recommendation systems, while also building interactive tools for content discovery that are used across the company.

As I progressed to Senior Data Scientist, I’ve taken on more complex projects involving real-time recommendation systems and multi-modal content understanding. Throughout my journey, I've pursued additional growth opportunities, including training to become a Bar Raiser (an Amazon program created to maintain high hiring standards through objective third-party interviewers).

Most recently, I've been exploring GenAI applications for internal productivity, developing AI tools for experiment write-up generation and interview preparation. 

How do you draw inspiration from culture and technology?

I follow developments in machine learning and AI closely, particularly tracking breakthroughs in natural language processing and generative AI through research publications and industry blogs. I get valuable exposure to cutting-edge work across the industry by attending external conferences like ACM Conference on Recommender Systems and ICML (International Conference on Machine Learning), as well as internal events such as the Amazon Machine Learning Conference. I've also taken courses on emerging topics like diffusion models and LLMOps (large language model operations) through Amazon, which has helped me identify opportunities to enhance both our internal productivity tools and customer-facing features. 

Can you share about a project you moved quickly to make something real?

I worked on a recommendation refinement feature that allows customers to narrow down book suggestions by selecting content attributes in real time. We developed a flexible framework with two components: one identifying relevant content attributes from a collection of books, and the other filtering recommendations based on customer selections. Collaboration was essential, so I worked closely with product managers, engineers and other scientists to define requirements, implement solutions and conduct thorough testing.

The outcome showed meaningful improvements in customer engagement and satisfaction during online testing, and it led to a full launch. We're now exploring ways to extend this approach to other discovery experiences.

Other than the people, what do you love most about the culture here?

I appreciate Audible's balance between innovation and impact. We're encouraged to experiment while having the resources to bring innovations to millions of customers. There's a genuine emphasis on ownership—I've been able to take projects from research through to implementation and see their real-world impact. Knowledge sharing is deeply embedded in our culture, with forums like research symposiums and tech talks creating opportunities to learn from colleagues across disciplines.

Perhaps most importantly, there's a sense that our work matters, whether it's helping customers discover their next favorite book or contributing to community initiatives.

What advice would you have for a candidate considering Audible?

Embrace Audible's entrepreneurial spirit. There's room to drive innovation and own meaningful projects. Consider how your skills might enhance the audio entertainment experience and be open to learning about this unique industry.

Build relationships across teams, as cross-functional collaboration is central to how we work. My most successful projects have involved partners from various disciplines.

Familiarize yourself with Audible's People Principles, as they genuinely guide our work.

Finally, bring your authentic self; diverse perspectives are valued here and create a more inclusive environment.

Lightning Round!

Favorite genre? Epic fantasy and sci-fi.

Listening from a speaker or headphones? Headphones.

Favorite activity while listening? Working out.

Morning person or night owl? Night owl.

Favorite snack? Popcorn.

What are you currently listening to?

I'm currently enjoying the litRPG series, Primal Hunter. Its captivating world-building introduces a dangerous multiverse governed by a mysterious System, and I’m invested in Jake's transformation from an ordinary office worker into a formidable hunter.

See open Global Insights & Data Science roles at Audible