Senior Applied Scientist (NLP/Search/RAG)

Share job

Senior Applied Scientist (NLP/Search/RAG)

Microsoft AI Organization aims to the best user experience for Web Search, Advertisement, Cloud, and Enterprise services. The Search Experience Group in Microsoft AI has more than 400 scientists and engineers, working on various NLP/Multi-modal techniques and applications.

We're looking for passionate and experienced engineers and scientists to help us on our mission of employing deep learning to understand all the data on the web - the largest store of information in human history. With this understanding we power end-user experiences across a variety of NLP/Multi-modal related areas, especially
- RAG System powered by LLMs
- Generative answers with LLMs
- Build up the world-class AI systems with novel NLP techniques and engineering excellence.
- Explore cutting-edge AI technology and deliver both research and production impact.

Required Qualifications:

- Experiences in applying deep learning, LLM, RAG techniques and drive E2E AI product development.
- 5+ years of working experience in NLP/search related areas.
- Passionate and self-motivated. Good communication skills, both verbal and written.

Preferred Qualifications:

- Master, advanced degree and/or industry experience

- Good experience in LLM post-train, finetuning

- Experience in RAG system

- Solid problem-solving skills, and ability to work independently

- Drive core technologies and E2E production delivery by leveraging State-of-Art AI technologies (especially LLMs).
- Address challenges in products through Deep Learning and Reinforcement Learning approaches and transfer novel ideas to production applications.
- Development of deep learning models for Microsoft AI scenarios, including generative search and answers, knowledge experience, et al.
- Pushing the envelope on deep learning by: (1) Defining problems and establishing metrics (2) Gathering training data at scale (3) Exploring model design and architecture (3) Exploring learning objectives and tasks