Guy Srinivasan works as a senior data scientist at Signifyd, but math is his first love. And his role at Signifyd gives him plenty of room to put his math background to use by coming up with new approaches to problems.
Srinivasan’s primary project is working on the machine learning model that assesses the fraud risk of ecommerce transactions for most of Signifyd’s customers. When a significant improvement to this model is ready for prime time, Signifyd minimizes merchant disruption during the transition by projecting which segments of traffic will be the most affected and making sure the roll-out is as smooth as possible for those segments in particular.
This becomes more complicated as the team scrutinizes more granular segments. Srinivasan has found that his math skills come in handy in this work.
“I think about the flow of information underlying a problem, which helps me see a different class of solutions – sometimes good, sometimes bad, but at least different, so worth considering,” he said. “One time I was implementing an algorithm to estimate the uncertainty in one of our projections, and I read a very nice summary on Wikipedia. But one step didn’t make intuitive sense. This nagging feeling prompted me to do some tests and I confirmed that the algorithm on Wikipedia (and thus copied all over the internet) was missing a factor.”
Exploring the intersection of machine learning and data science
Srinivasan’s path from a university math major — with plenty of computer science courses as well — to Signifyd’s data science team illustrates the flexible, innovative approach Signifyd takes to internal job moves.
After graduating from college, Srinivasan worked at Microsoft for eight years, then at Amazon for two years. At both companies, he was doing work he describes as “adjacent to machine learning.”
He then joined Signifyd as a software engineer — and his career journey since has allowed him to explore a number of paths. When he joined, Signifyd had 50 employees and everyone in Engineering worked as one team. As the company grew, they needed engineering managers, and Srinivasan was asked to become one.
“I learned that there are two kinds of functions: One is people management, and the other is team leadership and mentorship. I didn’t like any of the former, and I loved all of the latter,” Srinivasan said.
Fortunately, advancing into management in a single department is not the only way to grow at Signifyd which supports a dual-career track ladder. Srinivasan asked to move out of management, and he went back to software engineering. When Signifyd first built a data science team to work on automation and a machine learning framework, Srinivasan was on the software engineering team that worked closely with it: “We took everything data science was building and put it into production.”
Srinivasan was interested in data science – “I was one of the people who ranted about probability all the time” — and frequently engaged with that team. Almost three years ago, when the company wanted to shore up the engineering expertise on the data science team, Srinivasan was one of three engineers invited to join.
Facilitating this type of cross-departmental move has clear benefits for the employee who moves and helps the company retain employees who might otherwise look for new challenges elsewhere. But the benefits go beyond that, Senior HR Business Partner for R&D Trinity Malito said.
A chance to experience unexplored areas
“Supporting internal mobility creates a better workplace for everyone. While some benefits to the company are obvious, like reduced recruiting costs and faster onboarding, the wider impact includes renewed engagement, better retention and invaluable improvements to cross-collaboration and communication,” Malito said. “Employees transferring to a new team or department bring with them a wealth of organizational knowledge and connections. Employees can accelerate their careers through internal transfers because their knowledge and experience are often the key to solving the most challenging problems.”
Even while staying in the data science department, Srinivasan has been able to explore new types of work. “In data science, you can switch your day-to-day function without switching teams,” he said.
Since his move to data science, Srinivasan has changed focus or added responsibilities a few times, he said: “Even though my title didn’t change, I can go to my manager and say, ‘I’ve been focusing on roles B and C, but I’m really interested in understanding F. Can I have a project that puts me in the depths of F for this quarter?’”
His current primary project of working on a machine learning model, for example, was specifically chosen because he had been spending most of his time on features and tools for the last few years and wanted to gain experience with model building. It’s a high-priority feature that directly impacts Signifyd’s customers.
He is also working on a short-term project following up on one of the largest pain points members of the data science department have identified that could, if solved, make their jobs much easier and improve the operational excellence of the entire team.
“In addition to my usual responsibilities, I’ve been following up with individual data scientists who have run into major delays when running modeling experiments,” Srinivasan said. “I find out what specifically caused each piece of the delay, correlate the common causes, and work collaboratively with our aligned team in Engineering on low-hanging fruit mitigations.”
Signifyd encourages a roll-up-your-sleeves mentality
Indeed, Srinivasan has always felt free to explore ideas that are outside of his immediate assignment.
“I’ve always kind of assumed that if I see a problem somewhere, it’s fine for me to at least investigate it,” he said.
Underlying the culture of flexibility and development that makes these moves possible – both between departments and within one role —is a company philosophy that focuses on employee development.
“Career ladders that actually make sense are our greatest resource. While it can be a difficult resource to create, a truly great career ladder helps employees take a more active role in their personal development,” Malito said. “We combine this with regular manager training to move the career conversations further than just checking off boxes. We also offer a robust learning and development platform and stipend benefits for additional growth opportunities.”
In the future, Srinivasan hopes to continue building his career at Signifyd.
“I very much like the culture and the stability that it offers — the problems are great, and so are the people,” he said. Signifyd may at some point create a team dedicated to creating new fundamental capabilities for the various merchant-focused data science teams to use, he said — something that would align well with his interests, and could be a compelling next step in his journey at Signifyd.
Photo by Getty Images
Looking for a career that lets you grow and explore?