Data Scientist, Machine Learning - Fraud & Risk

Job Description

Come help us build the world's most reliable on-demand, logistics engine for delivery! You will join the team building the machine learning models to improve our fraud practices across our three audiences - Consumers, Dashers and Merchants.

As a machine learning data scientist focused on fraud & risk, you will lead the development of fraud detection algorithms and systems that protect DoorDash and its customers from fraud and financial loss.

You’ll be responsible for proactively identifying fraud patterns and sources to minimize the company’s exposure to financial and reputational risk. The ideal candidate has a high level of subject matter expertise and is experienced in utilizing information generated from fraud patterns, data analytics and business knowledge to identify insights and formulate a response plan. You will have the opportunity to leverage our robust data and machine learning infrastructure to develop models that protect millions of users across our three audiences.

The team is fast-paced, high-energy, and meticulous in diagnosing emerging fraud patterns and preventing fraud loss before it can happen. We conduct rigorous analysis of complex data sets, and sets up multiple layers of business rules, models, and other processes to prevent potential fraud. We are a highly cross-functional team and regularly engage in discussion and reviews with stakeholders to prioritize plans for reducing fraud impact and introducing safety features.

What You’ll Do

• Build and iterate on machine learning models to detect high risk activities like account takeovers, fake contents and fraudulent transactions
• Analyze effectiveness of existing fraud models and produce evaluations of the quality of detection algorithms
• Partner with Engineering, Product and Operations teams to conceive, design & monitor fraud risk strategies in order to mitigate fraud risk in the most precise manner possible
• Work with the business and product teams to develop, implement and maintain effective risk management programs to align with the business needs
• Identify and quantify levers to help move key metrics
• Recommend which product features to build and why
• Analyze rich user and transaction data to surface patterns, trends, and bugs that help inform our fraud policies and processes and contribute to our anti-fraud risk engine

About You

• High-energy and confident - you’ll do whatever it takes to win
• Detail Oriented - you do your due diligence when completing a task whether it’s one-off analysis, an ongoing investigation, or a broader project
• Get-it-done mindset. You’re not afraid of long hours, able to handle stress well, humble and scrappy!
• Adaptable, resilient, and able to thrive in ambiguity - things change quickly in our fast-paced startup and you’ll need to be able to keep up!
• Growth-minded - you’re eager to expand your skill set and excited to carve out your career path in a hyper-growth setting
• Desire for impact - ready to take on a lot of responsibility and work collaboratively with your team

Qualifications

• 4+ years of industry experience developing machine learning models in the fraud or risk space
• Experience developing and deploying machine learning / deep learning solutions
• Knowledge of fraud and abuse
• B.S., M.S., or PhD. in Statistics, Computer Science, Math, Physics, Economics, or other quantitative field
• Demonstrated programming abilities (Python preferred)

About Us

Founded in 2013, DoorDash is a San Francisco-based technology company passionate about transforming local businesses and dedicated to enabling new ways of working, earning, and living. Today, DoorDash connects customers with their favorite local and national restaurants in more than 1,000 cities across the United States and Canada. By building intelligent, last-mile delivery technology for local cities, DoorDash aims to connect people with the things they care about — one dash at a time.