Teresa Nieten

Data Scientist in the new Data & AI organization at FedEx. Using data science, including ML and AI, to streamline the clearance process for better, faster service.

At Ocean Insight, I combined my computer engineering experience with data science, turning light into insight. My primary focus was machine learning, where I transformed spectroscopic information into answers by developing a flexible framework that allows individuals without machine learning, spectroscopic or software engineering expertise to analyze their data using complex machine learning algorithms.

Supported the data analytic needs of the Veteran’s Benefits Association (VBA)’s Evaluation, Training, and Performance Improvement division. I evaluate results from post-training evaluation exams, post-training surveys, and job proficiency data to discover trends, using both aggregate and longitudinal analysis, all within R.

Employed a variety of data analysis tools including text analysis, predictive and forensic modeling, to discover patterns that could indicate customer dissatisfaction, fraudulent activity, or customer relationships. Examine customer feedback and activity for use in detecting or confirming fraud, using a combination of text analytics, statistical modeling, and classification; calculate sentiment score to assess client satisfaction and anticipate attrition; visualize customer relationships through graph modeling.

Research Scientist in the area of natural language programming in a SOA environment. A large part of this role was to research and apply theories, algorithms, and principals from Natural Language Processing (NLP), machine learning, and information fusion to formulate innovative solutions to the complex problem of finding actionable information in large corpora of unstructured text. Some of these algorithms were used to develop sophisticated NLP components to perform extraction of entities and events in context from natural language using a hybrid rule-based and a heuristics-driven approach.

Developed a framework for intelligent autonomous behaviors for unmanned system, on both simulation (OneSAF Test Bed) and live robotic platforms. The framework followed the Joint Architecture for Unmanned Systems (JAUS/AS-4) standard and contained a set of simple behaviors and a decision tree framework and editor to allow the creation of complex and unit-level behaviors using those simple behaviors.

Developed a framework for intelligent autonomous behaviors for unmanned system, on both simulation (OneSAF Test Bed) and live robotic platforms. The framework followed the Joint Architecture for Unmanned Systems (JAUS/AS-4) standard and contained a set of simple behaviors and a decision tree framework and editor to allow the creation of complex and unit-level behaviors using those simple behaviors. *The project team and software was transitioned to Interoptek in 2005