Jiayi Cox

The only hire in the whole company for designing and executing machine learning methods for scientific research.

Collaborated with the omics team to interrogate machine learning model output to deliver high value readouts.

Upgraded current in-house bioinformatics workflow by comparing and integrating several popular methods.

Reduced errors from time series prediction by proposing and implementing a working deep learning (LSTM) model.

Contributed a variety of functions to the company’s deep learning code base through Git and improved the performance of a vital deep neural network.

Identified the potential performance weaknesses of the current deep neural network in production, developed a concrete quality control framework for identifying results with high predictive error.

Made an entertaining Slack bot using Markov Chain to impersonate colleagues during a hackathon.

Applied various statistical modeling techniques on human high dimensional genomic data. Increased the productivity of research by developing an efficient HLA analysis pipeline, a linear mixed effect model pipeline and a data visualization pipeline.

Contributed to several potential publications by helping collaborators and providing key results.