Senior/staff AI Research Engineer

Hume AI Hume AI · AI Frontier · New York, NY +1 · Remote · Full-time

This role focuses on building and scaling state-of-the-art speech-language models (SLMs) using novel training methods like reinforcement learning from human expression (RLHE). The work involves supporting distributed model training, inference, and benchmarking, as well as large-scale data processing, with a goal of optimizing AI for human well-being and emotional expression.

What you'd actually do

  1. build out software systems that support distributed model training, inference, and benchmarking as well as massive-scale data collection, storage, preprocessing, and analysis
  2. working to solve some of today’s most exciting AI research problems at industry scale
  3. contribute to our endeavor to ensure that AI is guided by human values
  4. ensure that AI is guided by human values, the most pivotal challenge (and opportunity) of the 21st century
  5. contribute to our endeavor to ensure that AI is guided by human values

Skills

Required

  • Expertise in the Python ecosystem
  • popular ML libraries and tools (e.g. PyTorch)
  • writing robust and maintainable production-ready code
  • iterating quickly on new and uncertain research directions
  • 2+ years of experience training and/or fine-tuning transformer models with large-scale datasets of text, audio, image, and/or video data

What the JD emphasized

  • state-of-the-art speech-language models (SLMs)
  • reinforcement learning from human expression (RLHE)
  • human values
  • speech-language models
  • empathic AI assistants
  • human emotional expression

Other signals

  • building state-of-the-art speech-language models (SLMs)
  • reinforcement learning from human expression (RLHE)
  • AI is guided by human values
  • launch the next generation of Octave and EVI
  • foundational speech-language models for voice generation and powering real-time empathic AI assistants