Data Scientist II

The Applied Decision Science (ADS) team builds production ML models and risk management tools that are the core of Signifyd's product. We help businesses of all sizes minimize their fraud exposure and grow their sales. We improve the e-commerce shopping experience for everyone by reducing the friction experienced by good buyers and blocking fraudulent purchase attempts.

ADS builds and manages the entire decision stack - from designing and deploying the ML models that assess the riskiness of a transaction, to building the tools the Risk team uses to manage and fight fraud. We seek to standardize and automate repetitive work so we can spend more time on experiments and high-leverage projects.

We value collaboration and team ownership. Data scientists in Signifyd are true “full stack” operators, requiring knowledge of how transaction information received via our API traverses its way through our system and into the models we are responsible for building. When you test a hypothesis at Signifyd, you’re responsible for the end-to-end development, deployment, and evaluation process. This is a massive responsibility, and no one should feel like they're solving a hard problem alone. Together we help each other develop our skillsets through peer review of experiments and code, group paper study to deepen our machine learning and statistical understanding, and frequent knowledge-sharing through live demos, write-ups, and cross-team projects. All team members are expected and encouraged to weigh in as an external reviewer on a peer's idea or approach, regardless of level.

A couple quick notes on the Signifyd culture:

• We’re no stranger to remote work. Most of our workforce (ICs and leaders) are primarily remote. We tend to gather individual teams together once a year. There is no travel requirement for this role.
• We are heavy Slack users.
• We are heavy users of generative AI tools. We dislike token-maxxing, but enjoy the expansion of capabilities that have come with genAI. We ask that during the interview you don’t use genAI, as we want to know what you know.

Responsibilities: 

• Partner with the Business Unit Lead and their merchant portfolio to identify gaps in decisioning performance and implement solutions, with guidance from senior team members.
• Utilize existing, or build net new production machine learning models that identify fraud, in collaboration with other data scientists and machine learning engineers.
• Identify and build automation that reduces repetitive manual work.
• Run experiments to identify optimal decisioning strategies, balancing complexity and performance.
• Communicate complex ideas to a variety of audiences, from Customer Success and Sales, to limited interactions with external customers.
• Write production and offline analytical code in Python.
• Work with distributed data pipelines in Spark/Databricks/GCP.

Requirements: 

• A degree in computer science or a comparable analytical field.
• 3+ years of post-undergrad work experience required.
• Strong verbal and written communication skills.
• Strong machine learning and statistical background.
• Write code and review others' in a shared codebase in Python.
• Practical SQL knowledge.
• Design experiments and collect data.
• Experience with distributed analytics and data tooling such as Spark and Databricks.
• This role has on-call shifts, as part of our weekend rotation, Fri/Sat/Sun. While the number of shifts is subject to change, currently it works out to about six weekends a year.

Nice to Have:

• Previous work in fraud, payments, or e-commerce.
• Data analysis in a distributed environment.
• A passion for writing well-tested production-grade code.
• Experience with AI coding agents and automation.
• Experience of running A/B tests in production environments.
• An advanced degree.