Machine
Learning and Data Modeling – Model Productionization:
–
Facilitates
the implementation of machine learning (ML) model productionization processes
and process improvements.
–
Uses
technical knowledge and business familiarity to empower the transformation of
machine learning prototypes into production-ready models.
–
Implements
strategy to build technical expertise and readiness across team related to
model productionization.
–
Alleviates
the impact of obstacles on collaboration with multiple stakeholders, such as
Development Leads, Product Management, Operations, and Release Management, to
make, adopt, and communicate technical decisions, and shape the development and
delivery of software.
Model
Development and Deployment – Model Deployment:
–
Implements
multiple team standards around ML model readiness for deployment (e.g., model
scaling, model code cleaning, and meeting production quality standards).
–
Promotes
multiple team strategy around the automation of machine learning workflows,
from data extraction, transformation,
and loading (ETL) to model deployment and
monitoring, to establish the continuous integration and continuous delivery of
machine learning solutions.
Model
Development and Deployment – Model Performance:
–
Promotes
multiple team strategies around trained model/system alignment with design
criteria.
–
Identifies
improvements within multiple team processes around deployed model performance
evaluation and troubleshooting.
–
Facilitates
the creation of novel metrics that provide analytical insights to non-technical
stakeholders into how well machine learning models are operating.
Model
Development and Deployment – Data Quality:
–
Implements
improvements to multiple team processes for the identification and evaluation
of potential issues related to data quality (e.g., bias, fairness), data
security, and data privacy, and the minimization of their impacts on data
analyses and modeling.