Principal Data Scientist

Roche•Santa Clara, CA

1d•Onsite

About The Position

At Roche, we foster a culture where you can be yourself and are embraced for your unique qualities. Our culture encourages personal expression, open dialogue, and genuine connections, where you are valued, accepted, and respected for who you are, allowing you to thrive both personally and professionally. This is how we aim to prevent, stop, and cure diseases and ensure everyone has access to healthcare today and for generations to come. Join Roche, where every voice matters. Roche Sequencing Solutions is expanding with a Principal Data Scientist to lead and help drive forward R&D activity by providing a deep level of insight into experimental data, its interpretation, and implications for project development efforts. This role seeks a highly advanced and versatile individual who combines deep statistical expertise with proficiency in advanced data analysis tools and a focus on applying these skills to complex biological problems. (Applicants whose primary experience is narrowly defined within established bioinformatics analysis workflows might not align for the scope of this Principal Data Scientist role.) Principal data scientists combine a number of skills from different domains to organize, process, and learn from data, often through the lens of domain-expert informed models that help to abstract concepts from the data, test their validity, and make predictions.

Requirements

PhD in Statistics, Data Science, Computer Science, Engineering, or other related areas of study
5 years (or more) of experience handling various data sets/modules
Ability to define and lead analysis projects interfacing with multiple stakeholders
Demonstrated experience handling large datasets, with demonstrated ability of transferring the data into meaningful and actionable reports
Demonstrated experience in the application of statistical methods for process control and optimization (e.g., Statistical Process Control, A/B testing design/analysis, or advanced time series modeling) in a non-biological context
Demonstrated level of experience in a wide range of ML algorithms (traditional to advanced) and a strong understanding of the principles behind model training, validation and hyperparameter tuning
Practical experience in designing and implementing automated workflows; the ability to troubleshoot/map out a solution
Experience with genome sequencing data from multiple technologies
Experience using machine learning to solve biological problems
Proficient in various programming tools, e.g. Python, R, Java, C++

Nice To Haves

Experience with DOE for biological and assay development experiments, where we need to draw meaningful conclusions on small sample sizes.
Demonstrated level of proficiency with verbal and written communication skills.
Team player known for fanatical attention to detail at scale in a fast-paced environment.

Responsibilities

Analyze large datasets, identify patterns, correlations, and anomalies that might be hidden within the data.
Use statistical methods and machine learning algorithms to extract meaningful insights.
Provide statistical analysis of experimental results and communicate it to internal customers in a way that is both approachable and informative.
Work cross functionally using various pilot studies to then integrate the necessary analytical or inferential routines into a pipeline through a collaborative software engineering effort; the overall goal resulting in a validated internal software product that may be used regularly by experimentalists to track their research progress.
Utilize AI based tools for data analysis and create AI based support tools for internal customers.
Create clear and informative visualizations of the data and analysis results, making it easier for the technical (and non technical) team(s) to understand the situation and make informed decisions.
Identify technical challenges as they pertain to your team as well as collaborating teams, definite requirements and architecture for next-generation machine learning and statistical analysis products.