Skip to main content

Egra

AI Researcher / Engineer / Intern

New York City$125k–$300kfulltimeinternAdded 1 month ago

About this role

Join Egra as an AI researcher/engineer to develop multimodal foundation models trained on physiological signals, eye-tracking, video, and audio data at scale. You'll own projects end-to-end with immediate impact on the product, blending research and engineering with minimal process overhead.

What you'll do

  • Design self-supervised pretraining objectives for multimodal physiological and content data
  • Test foundation model robustness under distribution shift and identify failure modes
  • Build evaluation protocols to measure genuine progress versus benchmark artifacts
  • Develop internal research infrastructure for experiment tracking and dataset management
  • Bridge offline model research with live product performance
  • Document experiments and findings as internal knowledge base

What they're looking for

  • Machine learning research and implementation
  • Multimodal data handling and preprocessing
  • Model architecture design and evaluation
  • AI-assisted coding tools (Claude, Cursor, Codex)
  • Fast prototyping and iteration
  • Critical evaluation of benchmarks and representations
  • Self-directed problem-solving
  • Signal processing on heterogeneous data

Benefits

  • Complete ownership and autonomy from day one
  • Minimal onboarding or approval processes
  • Work deployed to production within weeks
  • Collaborative environment with co-founders
  • Flexible backgrounds welcome (PhD optional)
Apply with Autofill

Opens the application — the Jobs AI extension fills it for you. Set up autofill

Opens the official application on the employer’s site. No login required.

Egra

Egra develops multimodal foundation models trained on physiological signals, eye-tracking, video, and audio data. The company is hiring AI researchers and engineers to own end-to-end projects that blend research and engineering with direct product impact.

View all jobs at Egra

Likely interview questions

  • Walk us through a project where you took a vague research direction and shipped something concrete end-to-end. What was your timeline, and what did you have to figure out on your own?
  • Tell us about a time you looked at a benchmark or dataset and realized it was leaky or didn't measure what it claimed to. How did you catch it, and what did you do about it?
Unlock all 7 questions free — and practice them live →