Applied Scientist, Natural Language Generation (NLG)

Cupertino, California, United States
Software and Services

Summary

Posted:
Weekly Hours: 40
Role Number:200538807
Text generation is a key enabler for accelerated text input and intelligent interaction on Apple platforms. Our team is working on redefining user interaction with generative models for text generation. If you want to be part of an ambitious, organized and collaborative team that ships user experiences with pioneering ML applied to NLP, join the Input Experience NLP team in Software Engineering. Recent announcements at WWDC 2023 for running Transformer models on every keystroke for auto correction, sentence level correction and inline completions were conceived and built end-to-end on our team. You will work with a hard-working and dedicated set of outstanding ML and software engineers on a wide range of text generation technologies such as long-form text generation, summarization, question-answering, etc. We have been working in this area for years and own the NLP and ML text input stack for the keyboard input that includes auto correction, predictive typing on all Apple platforms. We also work on full stack ML applied to NLP and expose these key technologies across Apple on device and also to third party applications through the Natural-Language framework. If you want to amplify your strong ML and NLP skills into user experiences that will reach every person around you, this is the perfect opportunity! We exemplify Apple’s outstanding integration of hardware and software to create seamless input experiences. You will have the opportunity to go from building offline pioneering NLP models to optimization of the models for different hardware backends and user interfaces that make the experience magical. Our vision always includes a deep dedication to strengthening Apple’s privacy policy by achieving all of the above on device with powerful ML. Here are a selection of relevant WWDC presentations focusing on technologies: https://developer.apple.com/videos/play/wwdc2023/10042/ (https://developer.apple.com/videos/play/wwdc2023/10042/) https://developer.apple.com/videos/play/wwdc2020/10657/ (https://developer.apple.com/videos/play/wwdc2020/10657/) https://developer.apple.com/videos/play/wwdc2019/232/ (https://developer.apple.com/videos/play/wwdc2019/232/)

Key Qualifications

  • Experience building and then iteratively refining model pipelines end-to-end, from data curation to evaluation
  • Ability to design and perform experiments that bring ML and NLP research ideas to production
  • Familiarity with LLMs, such as fine tuning, prompt engineering, data synthesis, automatic evaluation, and retrieval-augmented generation
  • Proficiency in Python and ML toolkits such as PyTorch or TensorFlow
  • Excellent written and verbal communication skills
  • (Preferred) Background in linguistics, fluency in multiple languages, or a passion for scaling NLP-driven features for global audiences
  • (Preferred) History of publications, innovations, and/or leadership

Description

We are looking for a NLG Applied Scientist to innovate and enhance the input experience across all Apple platforms. If you are passionate about using groundbreaking ML for building phenomenal ML/NLP technologies, products and impacting user experience across billions of users, this is the job for you. The role will require you to set new directions, perform hands-on experimentation and implement your vision into deliverables. You will have wide impact across Apple with strong multi-functional collaboration across teams in hardware, software and design. Come join us! As a NLG applied scientist, you will research, design and develop machine learning models as well as tools/pipelines for input experience across all Apple platforms. Your role will have a direct impact on shaping the future of input technology roadmap. You will partner with data and evaluation teams to iterate on model quality, reasoning about both utilizing private user data and scaling to languages from across the globe. You will work closely with the engineering teams, perform hands-on experiments applying groundbreaking NLP and ML, derive insights from experiments and convert them into features that reach the hands of billions of users. KEY R&D PROJECTS PRODUCTIZED ON THE TEAM: Auto correction QuickType: Predictive typing Quickpath: Swipe input Image captioning Private Federated Learning for Language Models Natural Language APIs, NLP in CreateML

Education & Experience

PhD in Computer Science or a related field OR MS in Computer Science or related field with at least 3 years of industry experience

Additional Requirements

Pay & Benefits