# Applied Scientist 2

**Company**: Microsoft
**Location**: Beijing
**Experience**: mid
**Job type**: full-time
**Category**: Engineering
**Industry**: Technology
**Ticker**: MSFT
**Wikidata**: https://www.wikidata.org/wiki/Q2283

**Apply**: https://microsoft.ai/job/applied-scientist-2-5/
**Canonical**: https://yubhub.co/jobs/job_46016641-027

## Description

Are you interested in building personalized recommendations for billions of users, especially in finance domains? The Finance Recommendation team in Content Service organization is building personalized recommendations in finance domains in various products, including MSN and Edge default home page, etc. Our team focuses on the whole recommendation stack building, especially the modeling parts in different recommendation layers, including document understanding, segment recall, user profile modeling, personalized ranking, diversity optimization, etc.

As one applied scientist in the team, your major responsibilities include: Thinking through the product scenarios and goals, identifying key challenges and designing experimental processes for iteration and optimization. Offline model training and optimizing, collaborating with platform teams on online serving and perf optimization for shipping. Keeping on track of research trends in the fields of personalized recommendation, deep learning, and AI. Working independently and collaboratively with other team members.

Qualifications: Master/PhD in computer science or related fields focusing on machine learning and AI. 3+ years' experiences in areas of machine learning, natural language processing, or large-scale data mining. Solid coding skills and good experience on deep learning frameworks like PyTorch, TensorFlow, CNTK, etc. Preferred: Working/research experiences on recommendation areas are good plus. Good communication skills and self-motivation.

## Skills

### Required
- machine learning
- natural language processing
- large-scale data mining
- deep learning
- PyTorch
- TensorFlow
- CNTK

### Nice to have
- recommendation systems
- personalized recommendation