Embedded Data Scientist

You will develop and package AI models for execution on embedded microcontrollers and microprocessors used in handheld spectroscopic instruments. You will work closely with researchers and engineering teams to compress, characterize, and deploy models that enable real-world scientific and product outcomes. Onsite (5x a week) in Morrisville, NC.

Here's How You'll Make an Impact on the Team

Maintain and extend Python processes for generating and tuning training data from libraries of recorded instrument measurements

Train, characterize, and analyze model predictive performance against ground truth

Explore and implement model miniaturization strategies for embedded microcontrollers, balancing accuracy, precision, storage, and runtime

Optimize API utilization and identify configuration levers to support different model goals across hardware deployments

Assist with collection of instrument data for ML development and product characterization

Contribute to other company development tasks as needed

Maintain clear written and oral communication with stakeholders

Do you have experience in Version control?, Do you have a Bachelor's degree?, At least 2 years of experience with machine-learning frameworks such as TensorFlow and Keras

At least 2 years of experience in Python

At least 1 year of experience in C, Rust, or other embedded languages

At least 1 year of experience with Git or other revision-control systems

Bachelor's degree in Mathematics, Data Science, Computer Science, Computer Engineering, or a related discipline

Willingness and interest in learning specification, design, and manufacture of optical systems

Here's What Else Might Help You Out

Experience with LiteRT and STM32N6