Q&A with Olu?, Sr. Technical Product Manager at Amazon, US Prime & Marketing Tech, in Barcelona

Tell us about you and your career journey?

I’m Olu?, originally from Istanbul, Turkey and I have been living in various countries since 2010, most recently, Spain. I have an engineering background and I?previously led product teams in diverse industries such as hospitality, e-commerce, and on-demand delivery before joining Amazon in July 2022. At Amazon, I took on the role of Senior PM-T for the Discovery Tech team (DiscoTec), where I now lead automated content generation for cross-category shopping experiences for Amazon customers. I work closely with software development engineers, applied scientists and product marketing managers within our larger organization, US Prime & Marketing Tech team, Amazon page owners (e.g., Homepage, Search), personalization teams and business teams to deliver a cohesive customer experience worldwide on Amazon.com

?

Can you tell us more on the product you have been leading as a senior PMT?

To assist Amazon customers in their discovery and inspirational shopping journey year-round (e.g., shopping for Mother’s Day gifts or Summer Favorites), we launched curated product collections in the United States. This is an automated solution generating curated product selections that are high-quality, trending, tailored to top seasonal event themes and cater to customers’ shopping intents (e.g., Camping or Baking). Some examples are, Customers’ Most-Loved, surfacing products that are customer validated and highly rated, Seasonal Favorites (like Summer or Fall Favorites) surfacing products that are top picks for the season or gift selections for Mother’s Day or Father’s Day events. We use machine learning models to identify products for these themes and combine thousands of products from various categories through a streamlined automated solution. Customers can see our product collections through Amazon homepage, Search, or through our social media channels

?

What problem were you trying to solve and what challenges did you overcome along the product journey?

The primary challenge lies in understanding our customers’ pain points regarding discovery-focused shopping, similar to window-shopping. When Amazon customers are on a discovery-focused shopping journey, they often struggle to find a starting point, lack inspiring suggestions and feel overwhelmed by the multitude of options. While Amazon excels at serving customers who already have specific purchases in mind, there is still room for improvement for those who are ‘window shopping’ for fun or seek inspiration during their shopping journey. This is why my team’s favorite leadership principle is Customer Obsession. We aim to support Amazon customers on this journey by showing them exciting options they can consider purchasing and helping them in discovering more. We continuously question the kind of experience we want to create for our customers and strive to ensure that we are assisting them effectively. This mindset serves as a daily reminder for our team, which is why we always operate as if it’s “Day 1”.

?

What was the impact on customers and how did you measure success?

We observed significant customer engagement with our collections and received positive feedback through usability studies on how this solution helped them during their discovery journey. Customers appreciate having a starting point that inspires them during their cross-category shopping moments that are emotional, aspirational or recreational. We regularly ask our customers about their experience during major shopping events to see how these collections worked for them and where we can do better. In addition to qualitative feedback, we are checking key quantitative metrics focusing on customer engagement like clicks, purchase rates and returning customer rates. We also observe the metrics for customers who are new to Amazon or new to a certain category to understand if we are improving discoverability.

?

What technologies impacted your work?

To create product collections, we follow an expert-in-the-loop approach, where marketing expert feedback is augmented and amplified using historical data (e.g., what were customers searching and buying in this period last year?). We also use large language models (LLMs), retrieval augmented generation (to expand on key concepts from marketing feedback and retrieve items from our catalog) and distilled machine learning models (for fast and cost-effective inference). We use technologies such as Spark for data crunching, AWS Bedrock for working with LLMs and AWS Sagemaker for model lifecycle management. Finally, we have constant vigilance using Sagemaker GroundTruth to keep an eye out for drift and to continuously keep our models (and our pipelines) fresh.

?

What is the best part of working on this team and the career opportunities?

Our team believes that the “curated product collections” is an opportunity to address customer happiness and satisfaction. This means we are in the critical path for many of the broader Amazon retail goals, which enables us to be constantly challenged with bigger problems to solve, scale, and the ability to improve and iterate based on the latest/cutting edge solutions. Also, working on a problem that I can relate to as a customer has been the most rewarding aspect. Seeing the impact of our work firsthand and knowing we are making a positive difference has been incredibly fulfilling. Plus, collaborating with a diverse, multicultural team comprised of various job families to deliver these solutions to Amazon customers has been fantastic. Our teamwork is what makes our creations truly special – we learn and grow together every day.

?

If you are interested in joining Olu? within the US Prime & Marketing Technology team in the US or EMEA- check out all of our open roles here!

?