Data Scientist – Trust & Safety 109 views

Wish is a mobile e-commerce platform that flips traditional shopping on its head. They connect hundreds of millions of people with the widest selection of delightful, surprising, and—most importantly—affordable products delivered directly to their doors. Each day on Wish, millions of customers in more than 160 countries around the world discover new products. For over 1 million merchant partners, anyone with a good idea and a mobile phone can instantly tap into a global market.

They’re fueled by creating unique products and experiences that give people access to a new type of commerce, where all are welcome. If you’ve been searching for a supportive environment to chase your curiosity and use data to investigate the questions that matter most to you, this is the place.

Job Description

Wish has exciting opportunities for talented Data Scientists on their Trust & Safety Data Science team. This is a unique opportunity to conduct research, as well as develop and deploy Machine Learning models to address complex business questions such as fraud detection and anomaly detection at large scale.

Successful candidates will have extensive backgrounds in quantitative fields and significant experience with Machine Learning model development and deployment, as well as making impact with the end to end closed loop of data -> insights -> actions -> feedback.

In short, they are looking for people strong in: Machine learning theory and practice, data analytics and business acumen.

What You’ll Be Doing

• Working closely with the Chief Compliance Officer to write anti-money laundering rules to safeguard the business against expensive regulatory escalations.
• Maintain and approve machine learning models in one or both of these areas: machine learning models to detect misleading / ambiguous product listings on Wish and/or models to detect malicious product editing on Wish
• Working closely with a third party vendor and/or in-house anomaly detection tools to find fraud attacks on a large scale to defend the platform against large losses of money and user trust
• You will perform all aspects of data science in risk: data analytics, research, model development, deployment and monitoring – owning projects end to end, with concretely calculable impact.
• You identify the insights and data needed, and set priorities by the final impact.
• You understand the dual goals of the risk domain: fighting fraud and growing the business. You can clarify strategic and tactical objectives and prioritize tasks for yourself and the team.
• You proactively watch for signs of trouble, articulate with data, and craft hotfixes or long term solutions, depending on the situation and your judgement.
• You own foundational work that enables your direct colleagues as well as other teams.
• You enjoy managing stakeholders, multi-team projects, taking end-to-end responsibility and ownership over a project.

Qualifications

• Energetic and flexible. They own the actions. They own the impact. They iterate fast because fraudsters do.
• You are a team role model for technical excellence and communication style. You will provide input to technical decisions and uplevel the team’s ML capabilities in one or more major areas (e.g. misleading content models or malicious product edit models).
• Theory and practice of Machine Learning. 1+ years of hands-on industrial experience on most ML/DL algorithms and big data technologies.
• System design and programming. 1+ years of hands-on industrial experience with ML system design and implementation. They productionize code themselves, and they love it to be clean.

Preferred Qualifications

• Background in Payments, Risk or E-commerce strongly preferred
• Product sense. You always start from business problems, not technical details.
• Communication. You enjoy managing stakeholders, complex multi-team projects management, taking end-to-end responsibility and ownership over a project.
• Education. Masters in computer science, data science, or a related quantitative field.
• Domain expertise. Previous experience in fraud detection and the unique challenges applying ML in this area.