Just as Pinwheel, a payroll connectivity API company, was preparing to launch its Earnings Stream product last year, the company received an urgent request.
A client asked for a detailed expected coverage and accuracy analysis by the end of that week — a request that may have stopped some teams in their tracks. Yet Engineering Manager Andy Chen was ready to take on the challenge.
“Product Manager Devin Decaro-Brown and I decided to bootstrap the analysis ourselves, so we cobbled together various queries from previous analyses and delivered it on time,” he said.
Not only did this resourcefulness reflect the company’s value, “Underdog Mentality,” but it also highlighted the scrappiness and customer-centricity that went into building the Earnings Stream solution. Chen said the project — creating a product that delivers real-time customer income data — involved maintaining many moving parts, such as incorporating client feedback, fostering strong communication and, most importantly, working as one cohesive team.
Although Chen played a leading role in the product’s development, he believes his peers are the ones who made it all possible. Leveraging collaboration, passion and expertise, team members from across the engineering, product and data departments brought the solution to life.
Chen considers these efforts a reflection of the innovative spirit that has always defined the company.
“Pinwheel is a place where you can make anything happen if you’re willing to put in the time and effort to do so,” he said.
Built In interviewed Chen to find out what led to the product’s creation, how the task was carried out successfully and what sets Pinwheel apart from others in its industry.
IN PURSUIT OF FINANCIAL EQUALITY
Pinwheel is on a mission to build a fairer financial system. The company’s software enables businesses to gain insights into their customer’s earnings, simplify the process of retrieving tax forms, manage lending risk and more.
In conversations with customers, we kept hearing about a need for a comprehensive, real-time view into an end user’s earnings. Most of these conversations focused on unlocking the Earned Wage Access use case, but we realized this product could also be applied to many other use cases, such as spend management, credit limits and overdraft and default protection.
Since my team had access to both payroll data and domain expertise to derive insights, we were well positioned to build such a product. We now have a powerful Earnings Stream product that serves as the backbone of our Earned Wage Access implementation.
When we launched with our first Earnings Stream customer to power their Earned Wage Access use case, it had an immediate and tangible positive impact on their end users. One of their end user’s said, “Getting paid early gives my family greater flexibility for bills and spending without negatively impacting our finances.” Hearing this was an awesome moment of validation for us, because it proved that we had built a great product.
What role did you play in developing and launching the Earnings Stream product?
I wore many different hats for this project. I played the role of manager, engineer, data analyst, product manager and even customer support specialist. Ultimately, it was an extremely rewarding experience to lead my team in true end-to-end fashion to build this product.
A STRONG TECHNOLOGICAL BACKBONE
Chen said Pinwheel’s APIs are powered by a Python stack, FastAPI, PostgreSQL and Amazon Web Services. For data analysis, the company leverages Amazon Redshift, Looker, Jupyter Notebook and Hex.
What obstacles did you encounter along the way, and how did you successfully overcome them?
One obstacle we faced was that we knew what to build, but we didn’t know what form it should take. I remember many passionate debates between the tech lead, Emad Khan, and the product manager, Alex Wendland, on the ergonomics of the API. I credit both of them for going the extra mile to ensure that we were building the right product. Through their hard work, they arrived at the foundations for what the Earnings Stream API looks like today.
When this project began, I had just transitioned from software engineer to engineering manager. Through sheer instinct and habit, I leaned into my individual contributor skills a bit too much early on in the project. Eventually, my direct reports expressed that they felt they lacked independence and the ability to learn and grow with me being so much in the weeds and guiding them on every implementation detail.
After some self-reflection, I realized I needed to change my approach in order for the team and project to succeed long term. As a result, I took a step back from the implementation details and tried to be as hands-off as possible, set goals and expectations with my team and set up processes for regular check-ins. After making this change, my team felt happier and more productive, and I felt more effective with my time.
What teams did you collaborate with in order to get the product across the finish line, and what strategies did you employ to ensure smooth cross-functional collaboration?
The data team — which includes Margaret Turzanska, Bobby Mareddy and Ke Jiang — built the model that powers all of Earnings Stream. To ensure our teams were aligned, we embedded one of the engineers on my team, Steven Dias, into the data team. This ensured the data team had someone with product context, while our team gained someone with a deep understanding of the backing model. This strategy ended up working well, speeding up the time taken to productionize the model and validate its accuracy.
The product team provided a clear product strategy on what features were critical for launch and which ones could be added later. One feature that was critical was “unavailable reasons,” which was a set of rules that communicated why Earnings Stream data was unavailable. One of our senior engineers, Terren Suydam, built a high-quality feature despite tight deadlines. He did so by aligning with the product manager on edge cases, example scenarios and business logic up front to ensure that the actual implementation would be carried out quickly.
Throughout the project, we used several communication strategies to ensure stakeholders were aligned. This included dedicated Slack channels for specific features, weekly project updates on progress and blockers and a master launch checklist to ensure nothing slipped through the cracks.
How does Pinwheel compare to other companies in your industry when it comes to building and launching new products?
We place an emphasis on getting a version of our product into the hands of customers early so that we can get feedback more quickly. With Earnings Stream, we sent an implementation guide to a customer to get initial feedback before we built anything. Then, we built a sandbox version of the product with test data so that they could play with endpoints and webhooks to provide feedback on specific details. We were building the real production version while continuously incorporating their feedback.
We build as a team rather than as individuals. What matters most is that we build the right product rather than prioritizing who builds it.
“We build as a team rather than as individuals.”
For example, despite my team being newly formed, the company trusted us with developing Earnings Stream. I also had a strong tech lead on the team, Pranav Kunapuli, who acted as a stabilizing force. Ultimately, we succeeded because we were passionate and dedicated to building the best product.
There are countless instances that prove you can make things happen here, whether that involves learning new skills or transitioning into a new role. At Pinwheel, if there’s a will, there’s a way.