The Data Science team currently focuses on payments and retails. Every data scientist is expected to own an end-to-end lifecycle of framing business problem statements, building MVPs, designing models and systems, preparing data pipelines, deploying to production. Each Data Scientist typically focuses on 1 business domain at a time, with rotation opportunities and exposure to peers’ work. The team is looking for passionate data scientists who aim to become thinkers, leaders and pioneers. We work remotely from 3 sites: Ha Noi, Ho Chi Minh and Singapore.
- Develop a deep behavioral understanding and intuition of our users from data to identify emerging fraud trends, develop, and improve machine learning models to detect risk and fraud.
- Find suspicious trends from massive and rich sources of payment and retail data, identify gaps in current detection, communicate actionable insights and get feedback from risk operations and management, form strategies and action plans.
- Build hypotheses and reframe trends and insights into machine learning problem statements, implement and deploy models to detect fraud suspects.
- Build metrics, continuously monitor model performance, build automated and manual alerts on potential issues such as false positives, drastic model performance changes, pipeline failures.
- Think out of the box and innovate in all possible perspectives.
Minimum qualification (Exact responsibilities vary with levels)
- Postgraduate degrees in Machine Learning, Statistics, Applied Mathematics, Physics, Computer Science, Electrical/Computer Engineering, Industrial & Systems Engineering, or related technical field/experience.
- Proficient in Python and PySpark, or display ability to catch up quickly on Python. Proficient in SQL, or display ability to catch up in sequel.
- Understand and display ability to apply at least one of the following: Experimental Design, Linear Models, Multivariate Analysis, Stochastic Models, Sampling Methods, Classification Models, Clustering, Anomaly Detection, Graph-based Models.
- Display ability to: articulate business questions, reframe business questions into mathematical problems, set success metrics, identify data required, select the right models or use statistical techniques to arrive at answers using available data.
- Display ability to design and execute on prototype models.
- Responsible and self-motivated, curious and independent learner, constantly strive for improvement and enjoy sharing knowledge with team members.
- Detail-oriented and efficient time management in a dynamic and fast-paced working environment.
- Passionate about fraud detection
- Real-world implementation experience in machine learning including classification, clustering, and anomaly detection. Deep understanding and implementation experience of predictive modeling algorithms such as logistic regression, neural networks, forward propagation, decision trees and heuristic models, with familiarity dealing with trade-offs.
- Real-world experience in ETL, feature selections, hyper-parameter optimization, model validation and visualization.
- Experience in tools like Scikit-Learn, Pandas, and XGBoost.
- Experience in deep learning frameworks like Tensorflow or PyTorch.
- Experience researching or working in at least one of the following sectors: retails or e-commerce, financial or payment services, telecommunications, POS, social networks, travels, or other sectors within the VNLife ecosystem.
- Experience implementing research papers in code.
- Experience in interfacing with other teams and departments to deliver impact for the organization.