The posting
We're looking for an AI Engineer to help build the next generation of AI models behind our diagnostic platform. You'll work on modern, large-scale machine learning for health, turning large, messy, real-world datasets into models that carry real clinical signal.
You'll join a team of AI engineers who care about doing this the right way. We're looking for someone who brings current knowledge of the field, whether from graduate research or hands-on work, and who pushes the team to approach problems from a modern AI perspective. We want someone who understands why methods work, not someone who can only run a tutorial notebook on an off-the-shelf model. Together you'll assemble and curate large datasets, train and evaluate models, and show whether they carry real clinical signal.
Learning goes both ways here. You'll work alongside experienced AI engineers who review your code and talk through your results, and we expect you to teach them what's new. The work is hands-on from week one. Over time you'll own larger pieces of the pipeline, from data to deployed model.
Responsibilities
- Train and evaluate modern deep learning models on large datasets, including self-supervised approaches and fine-tuning of large pretrained models.
- Build and maintain the data pipelines that feed them: collecting, cleaning, deduplicating, and versioning large video datasets.
- Develop ways to learn from noisy or limited labels, and measure how much real signal the models carry.
- Run experiments carefully: track runs, make results reproducible, and write up what you found, including negative results.
- Design evaluations that reflect real-world use, and show clearly where offline metrics and product performance diverge.
- Support our existing computer vision work as models move toward production.
- Write clean, tested, well-documented Python that teammates can pick up and reuse.
- Take part in code review and technical discussions, and explain your results to both technical and non-technical colleagues.
- Stay on top of current AI research. Reproduce promising methods, propose experiments, and teach the team what you learn.
- Pitch in on whatever the team needs. We're small, and everyone does a bit of everything.
About Senseye
Senseye is a NeuroTechnology company based in Austin, TX, on the cusp of revolutionizing mental health. We're developing the world's first objective mental health diagnostics platform, measuring cognitive activity through the eye using mobile devices. Our initial diagnostic for PTSD is entering clinical trials, with additional indications for anxiety and depression to follow. We're a team of scientists, engineers, and clinicians building technology that could redefine how mental health is understood and treated.
Requirements
- Master's degree or PhD in computer science, engineering, mathematics, or a related field, or equivalent experience showing comparable machine learning skills, such as several years of applied ML work or substantial published or open-source work.
- Recent graduates are welcome. Up to about 3 years of hands-on ML experience, including graduate research, a thesis, internships, and industry work, is a plus.
- Solid Python and experience with at least one deep learning framework.
- A working grasp of ML fundamentals: training and validation, overfitting, evaluation metrics, and why a good test score isn't the same as a working product.
- Hands-on experience with modern deep learning beyond off-the-shelf use. You've trained or substantially adapted models yourself and can explain your design choices and the tradeoffs.
- Comfort with Git, the command line, and working on a shared codebase.
- Clear writing and communication. You can explain a result or a plot simply and say what you're unsure of.
- Curiosity and the habit of digging in when something looks off, in the data or in your own code.
Extra Points
None of these are required.
- Hands-on experience with self-supervised, contrastive, or masked-modeling methods.
- Experience with large-scale training: multi-GPU runs or experiment tracking such as Weights & Biases.
- Experience with video, time-series, or sensor data.
- Experience with noisy-label learning or semi-supervised methods.
- A paper, open-source contribution, or substantial personal project you can show us.
- Exposure to healthcare, medical device software, or other regulated settings.
Benefits
- Flexible schedule and flexible paid time off
- Health, dental, and vision insurance
- Short- and long-term disability and life insurance
- 401(k)
- Continuing education opportunities
- Health and Wellness stipend
- Fully stocked kitchen and a flexible hybrid environment
- The freedom and trust to shape your role as we design, build, and ship our products
