Evie Mahsem

Contractor, Midcontinent Independent System Operator (MISO) Developed analytical tools and systems to support fraud detection in MISO energy markets. • Developed and experimented with machine learning models to predict market participant behavior. • Created advanced queries to assemble complex datasets from disparate Oracle and Netezza databases into a single analysis-ready dataset. • Built interactive fraud-detection visualizations using Python (Plotly, Dash), designed for integration into team dashboards. • Collaborated closely with teammates to interpret complex datasets with little to no existing documentation, ensuring accurate and reliable analysis. • Initiated shared documentation to formalize data definitions and reduce future ambiguity.

1. Used Long Short Term Memory ML with Client ticket information from Service Now and Snowflake and utilized the ticket title to determine the appropriate support group to send it to. 2. Collaborated with a team in developing a predictive model that could predict project slippage, which could potentially save the client millions. 3. Utilized Topic Modeling on the IT survey data and developed a Power BI Report to help the IT team comprehend the content of their extensive response dataset.

Learned aspects of Big Data and applied them to coding projects individually and with a team.

Contributed to physics code development for CERN, in Python, which was maintained using GitHub. Of these projects, the most notable are: • To distinguish between signal and background in a vast quantity of data obtained from the Large Hadron Collider, the process of hyper-parameter tuning is employed for a Deep Neural Network (DNN). • Utilized C++ to generate plots that aid in determining the dimuon mass within the ROOT data analysis framework. • Experience reading particle physics textbooks and papers and applying the concepts using Python and C++.