Founding Machine Learning Engineer

About the Role

You will architect the machine learning stack for emotion inference, defining how raw neural signals become high-fidelity models. Your work will shape the core architecture of Orbit's foundational emotion models, directly influencing how wearable brain-computer interfaces translate neural activity into insights. You will own the journey from initial data streams to validated, privacy-aware products. This position requires deep collaboration with neuroscience, hardware, and software teams to ensure that data pipelines, model training, and deployment strategies function as a unified system.

Key Responsibilities

You will design and implement data ingestion frameworks that convert complex neural streams into structured, high-quality training datasets. Your architectural decisions will govern how PyTorch, TensorFlow, or JAX handle the unique constraints of wearable sensing. You will establish review protocols where quantitative emotion metrics directly inform model iteration and design choices. You will be responsible for translating experimental findings into robust, shippable features that adhere to strict privacy requirements. A critical part of your role will involve creating integrations that allow Orbit's models to interact with external platforms and clinical tools. You will set and maintain standards for data labeling that ensure consistency across multimodal inputs. You will pioneer rapid prototyping methods for fusing time-series and imaging data. You will also define deployment strategies that preserve model performance across edge devices and cloud infrastructure.

Qualifications

You hold a Bachelor's degree or higher in a quantitative field such as Computer Science, Electrical Engineering, Applied Mathematics, or another STEM discipline. Exceptional candidates without formal degrees will be considered based on demonstrable expertise. You bring multiple years of hands-on experience building and refining machine learning models, evidenced by work history, open-source contributions, or research publications. You possess advanced Python skills and have productionized models using major frameworks like PyTorch, TensorFlow, or JAX. Your background includes developing and refining machine learning pipelines that handle diverse and complex data types. You are experienced in preprocessing, labeling, and exploratory analysis of high-dimensional datasets. You work effectively with multimodal data, such as combining imaging with time-series signals or text with audio. You have a proven track record of succeeding in fast-paced, independent team environments.

Additional Assets

Publications in leading machine learning or specialized neuroscience conferences and journals are a strong asset. Experience with biomedical, neuroimaging, or other high-dimensional sensor data is highly relevant. A background in signal processing for time-series or imaging data is valuable. Familiarity with distributed or large-scale training methods, such as mixed precision across massive datasets, is a plus. Knowledge of semi-supervised or self-supervised learning techniques is beneficial. You are eager to immerse yourself in neuroscience and neuroimaging context to enhance your contributions.

About the Team

You will join a founding team responsible for building a generational NeuroAI stack. The team operates at the intersection of neuroscience, hardware, and software, tackling problems with significant autonomy and intensity. The environment demands fast learning and the ability to navigate open-ended challenges. Your work will directly contribute to redefining how technology interfaces with human emotion and wellbeing.

Compensation and Location

This is a full-time position based in San Francisco. The compensation package ranges from $180,000 to $360,000 annually, reflecting the level of impact and responsibility.

Practical Information

Please confirm specific details regarding the application process on the official page. We are backed by leading figures from AI, neurotech, consumer hardware, and pharmaceutical companies, and we operate as a venture-funded entity.

About the company

We're a team of engineers, neuroscientists, and designers solving the most difficult and meaningful challenge: understanding the human brain. Our translational brain computer interface and pioneering models decode emotion, putting experience and wellbeing at the center of every interaction.