Applied Scientist II, Financial Insights and Actions
Amazon · Vancouver, Canada · Research Science
About this role
Amazon is hiring a mid-level Applied Scientist in the machine learning function based in Vancouver, Canada. The posting calls out experience with Python, Java, LLMs, Distributed Systems. Compensation is listed at C$149,300–C$249,300 per year.
- Role
- Applied Scientist
- Function
- machine learning
- Level
- mid
- Track
- Individual contributor
- Employment
- Full-time
- Location
- Vancouver, Canada
- Department
- Research Science
- Posted
- Apr 10, 2026
More roles at Amazon
Job description
from Amazon careersAre you interested in changing the way accounting and finance works at Amazon? We are a science and engineering team leveraging ML models and GenAI/LLMs to solve real-world problems faced by accountants and financial analysts. We are part of the Amazon Financials Foundation Services (AFFS) organization. AFFS is responsible for processing and managing billions of financially relevant transactions sent globally from across Amazon each day, including orders, shipments, payments, and inventory movements. AFFS is at the center of Amazon's key initiatives and fuels the growth of Amazon's businesses worldwide by ensuring that businesses can easily integrate with our services and that accountants and financial analysts have the right tools to use our data. As an Applied Scientist, you'll work alongside domain experts, engineers, and other scientists to understand business problems, propose scientific solutions, and deploy them to production. You'll work on scientific initiatives for accelerating reconciliation, standardization, and onboarding. This includes: - Leveraging GenAI/LLMs to build agentic solutions to accelerate accounting-related research/tasks and produce proactive insights. - Building AI trust and safety in the financial domain. - Establishing scalable, efficient, automated processes for large-scale data analysis, machine learning model development, model validation, and serving. - Developing training/evaluation datasets for model…