October 18, 2022

We are looking for outstanding Ph.D. students interested in full-time scientist positions or 2023 spring/summer research internships (full-time or part-time) on the Human-in-the-loop Services (HIL) team.

The mission of Human-in-the-Loop services (HIL) is to accelerate the development of AI applications and solutions with human intelligence as an integral part for customers, ranging from tech-savvy ML engineers, data scientists, data operations managers, and program managers. We deliver this today through two main product offerings: SageMaker Ground Truth (GT) and Amazon Augmented Artificial Intelligence (A2I). GT enables customers to acquire high-quality labels and annotations for their machine-learning training datasets. A2I enables customers to build workflows that incorporate human judgment for production ML systems in which human monitoring or auditing are critical business requirements.

Amazon science in the HIL team spans the fields of machine learning, computer vision, and natural language processing, crowdsourcing/human computation. We fundamentally believe that scientific innovation is essential (https://www.amazon.science/about [amazon.science]). We have been working on topics including self-supervised learning for autonomous driving, data-centric ML (active learning, controllable text generation with structured causal models), joint vision and language representation learning, and foundation models.

For full-time applied scientist positions, please apply via
www.amazon.jobs/jobs/1926529?no_int_redir=1 [amazon.jobs]
www.amazon.jobs/jobs/2038547?no_int_redir=1 [amazon.jobs]

Each intern will work on a 3-month project with our scientists. We encourage our interns to publish their results. Interested Ph.D. students are encouraged to apply via
www.amazon.jobs/jobs/2156173?no_int_redir=1 [amazon.jobs] for machine learning,
www.amazon.jobs/jobs/2156148?no_int_redir=1 [amazon.jobs] for computer vision,
www.amazon.jobs/jobs/2156175?no_int_redir=1 [amazon.jobs] for NLP
AND to send their resume directly to aws-a2i-science-interns@amazon.com.

We are actively reviewing applicants now and positions will close once filled. Please share this announcement with students in your network or apply directly.


Li Erran Li (Amazon), Stefano Ermon (Amazon Scholar and Stanford) and Pietro Perona (Amazon Fellow and Caltech)
Amazon Human-in-the-loop Services

Some of our recent papers:
Self-Supervised Pretraining for Large-Scale Point Clouds [nips.cc], NeurIPS’22
Neural Attentive Circuits [nips.cc], NeurIPS’22
A Causal Lens for Controllable Text Generation [proceedings.neurips.cc], NeurIPS’21
Humble Teachers Teach Better Students for Semi-supervised Object Detection [openaccess.thecvf.com], CVPR’21