Deepak Agarwal: My First 90 Days at Pinterest
Deepak Agarwal | Head of Content & Discovery
Changing jobs in the middle of a global pandemic? Stressful. Joining a company to lead Engineering for a product that millions look to for inspiration during said global pandemic — doubly so, but also incredibly rewarding. We sat down with Deepak Agarwal, Pinterest’s new Head of Content & Discovery, to discuss his first three months at Pinterest and his vision for the team’s future.
Let’s start with some background. What work are you most proud of thus far in your career?
I joined Yahoo! as an individual contributor when the site’s homepage was the most visited site on the internet. It was entirely hand-curated, though, and our goal was to automate it through machine learning. This work had a tremendous impact on Yahoo’s business, and great technology was created in the process. Back in those days, the Yahoo! front page was the most visited site in online media. It was a firehose that shaped traffic to other downstream Yahoo! pages like Finance and Sports, and to many other external media sites like TV network pages, etc., that were dependent on Yahoo! for leads and traffic. This ML work led to a 200% increase in digital traffic across the board.
After leaving Yahoo!, I joined LinkedIn, and my priorities shifted in that I was now responsible for developing employees and catalyzing the success of others. We built a team that infused intelligence into everything that happened on LinkedIn. Data is the foundation of LinkedIn’s business, and everything across the ecosystem is based on recommendations and search both across consumer and enterprise (e.g.,include feed, people recommendations, advertising, search, jobs recommendations, recruiter and sales). In addition, keeping the site safe and trustworthy is of paramount importance. The team built Ml technology that made every single aspect above more intelligent at scale via sophisticated algorithms.
I was lucky to have the opportunity to develop great leaders and work with some of the brightest minds in the industry. Our team had a significant impact on LinkedIn’s product and culture, and machine learning went from a nice-to-have to a must have under my leadership. I am grateful for these experiences, and thrilled that they have led me to my current position at Pinterest.
Throughout the last few months, what has stood out to you about Pinterest’s Engineering organization?
Pinterest has markedly sophisticated tech both currently rolled out and in production. Our personalized page rank and computer vision are cutting edge, and our automated machine learning can train complex, deep learning models reliably and at scale. The teams supporting this tech are small and nimble, and we have some of the most talented engineers in the business. I am continually impressed by the work I see and the scale at which it’s deployed.
As a machine learning leader, what drew you to Pinterest?
The Pinterest product is already machine learning-first, which is a huge benefit for ML practitioners who want to do cutting edge work. Content is the core of our product, and it would be impossible to surface the right content to the right Pinners without machine learning. Pinners love Pinterest because we have built sophisticated models that recommend the right idea at the right time, then curate additional cross-vertical ideas from that point. We are using innovative technologies every day: automated intelligence, computer vision, augmented reality, 3D animations, recommendation and search and e-commerce, all in the service of building a great product for Pinners.
For someone considering joining Pinterest, how would you describe our technical opportunities?
Our challenges, and therefore our opportunities, are fundamentally about the psychology of our Pinners. Because our product was created to “bring everyone the inspiration to create a life they love,” in order to live our mission we must go far deeper than simple surface level recommendations, and get to the fundamentals of our Pinners’ tastes. For example, if a Pinner searched “avocado” on Pinterest, we would take it one step further and recommend Guacamole recipes, creative takes on avocado toast, and even pestle and mortar merchants. Pinterest sits at the intersection of machine learning and consumer intent, and that opportunity is one of the reasons I joined the team.
Another major benefit of working at Pinterest is our data. People have saved more than 240 billion Pins, and the service is used by 416+ million people every month, discovering and curating this content to their boards. This ongoing cycle of data powers a sophisticated recommendation and search system, and our content organization is unique and produces strong engagement. Pinterest is known as one of the only places where people actually want to see ads, since they help facilitate the conversion from inspiration to action and make Pinner’s goals and dreams more tangible.
The next frontier at Pinterest is taking Pinner inspiration, and translating it into confident action across a variety of verticals. For example, if a Pinner comes to Pinterest searching for home decor, we want to make it easy to purchase the products they discover (a new rug, lampshade, sofa), and guide their search to similar items across art, fashion, or architecture. Our teams are actively building the technology that powers this.
At the highest level, Pinterest’s team of engineers is working to bring inspiration to hundreds of millions of people. In normal times, this is a great opportunity and responsibility. During these unprecedented and challenging times, the impact is profound.
You’ve touched on it a bit, but what specific technologies are your teams building out these days?
Another aspect of Pinterest that I appreciate is that we are never building tech for the sake of tech. Every single project we embark on is tied directly to the goal of improving the Pinner experience. Some projects we’re currently working on:
- We’re creating cutting edge inclusive search and recommendations with our new visual skin tone model
- We’re building complex, deep learning AutoML models that can deploy themselves in production and power large scale recommendations
- We’re building highly sophisticated neural networks
- We’re exploring Computer Vision, 3D, Augmented Reality, visual search, and shopping to create new experiences in-product
- We’re continuing to enhance our home feed which serves as a key touchpoint in a Pinner’s journey from inspiration to action
- We’ve created the Today tab as a source of daily inspiration with curated topics and trending pins
What is the biggest opportunity you see for Content and Discovery in Pinterest’s future? What work are you most excited to see in the next year?
A new suite of features for creators will unleash a brand new content source that is fully native to Pinterest. With this new corpus, we’ll be able to distribute content in a way that delights Pinners and encourages creators to keep creating — taking inspiration to a whole new level.
Content and Discovery is partnering closely with the Creators team on this work. We’re aiming to connect Creators with the right audiences to increase engagement and deepen the mutually beneficial relationship between Pinterest and our Creator cohort. The more positive feedback Creators receive in the form of content engagement (likes, shares, etc.), the more likely they are to continue creating.
Shopping is another frontier we’re exploring to truly convert inspiration to confident action. We’re prioritizing merchants on the home feed as well as in recommendation and search to keep transactions top-of-mind for Pinners in a natural way that caters to their taste and preferences.
Pinterest’s culture is evolving every day, and we are working hard to create a place where our employees feel comfortable and empowered to do their best work. As a leader and culture carrier, what do you see for Pinterest’s way forward?
With a strong and thriving workplace culture, employees come to work every day (virtually nowadays) feeling psychologically safe. In my experience, the greatest innovations are made when employees are unafraid, fail-forward, and open to asking for help. Building a culture like this does not happen in one day, and cannot be done by one person. All of us, especially at the leadership level, are responsible for creating this type of environment.
Based on my impressions three months in, there are so many great things happening culturally at Pinterest: our collaboration and productivity are both best-in-class. Pinterest is unique in that there are nimble teams doing very big things. In fact, Each engineer is responsible for hundreds of thousands of Pinners. Success is only possible when teams have a shared vision and mission, and everyone is fully bought in.
As a company that recently went public, we’re in a growth phase across so many areas — technically, with development constantly scaling; staffing, as we’re rapidly ramping up resources; and culturally, continuing to learn and improve our workplace for everyone. The intentional leadership we’ve seen from Ben (CEO and Co-founder) on down is so promising and I’m glad to be a part of it.
Do you have any closing thoughts?
Overall, leading Content & Discovery at Pinterest has been an inspiring opportunity for me. I am excited to continue building great tech in the service of enabling our Pinners to create lives they love.