Price Negotiation Model Development-Scraped data from various sources and conducted data cleaning and preprocessing tasks using Python libraries such as: Beautiful Soup, Scrapy, Pandas, NumPy, and/or Regular Expressions.-Developed a custom pipeline to predict product prices based on their specifications or attributes, and conducted statisticalanalysis to identify product margins and derive insights about market trends and competition.-Designed and implemented a custom clustering algorithm from scratch to group similar products based on their attributes.-Utilized recursive feature elimination with random forest for efficient feature selection, enhancing model performance andlowering computational complexity in the price prediction pipelinePredictive Modeling -Developed proof of concept models to predict cost for several clients across various industries-Utilized various modeling architectures to both classify costs and predict costs at a point in time-Experienced in cleaning large data sets, EDA, advanced training/evaluation techniques and data visualization-Created, managed and git tracked; data cleaning, modeling and evaluation pipeline in PythonApplied hyperparameter optimization using cross validated grid search, significantly improving in model performanceDigital Transformation-Conducted dozens of client interviews to develop an accurate and complete current state of the digital ecosystem-Completed comparative analyses between client and competitor digital ecosystems in order to create a set of actionableobjectives, correcting pain points in the ecosystem-Compared data warehouses like Snowflake, Redshift, and BigQuery to determine the optimal solution, enhancing data storage and query performance for the digital transformation project-Completed a roadmap of objectives grouped by development epic and assigned to objective owner

-Assisted hundreds of students weekly with 3D printing and 2D laser cutting-Certified hundreds of students for lab access-Repaired 3D printers and laser printers

-Worked directly under the CTO to create a machine learning models using 500,000 unique and cleaned VanEck.com web visitsto classify web users by investor type and identify bots-Predicted a user’s investor type based on their self-identification by creating and optimizing MLP and SVC classifiers-Defined and predicted a natural label for users by utilizing scikit-learn to create and optimize NMF, K-Means and-DBSCAN models that cluster users based on similarities of their web sessions-Identified bots by plotting users based on median theta deviation and MAD theta deviation from their mean web session

-Directed full stack app development team for the non-profit Cultivando Juntos, focusing on the nutritional needs of underprivileged farm workers-Created features in React.js to host news, host videos and keep track of various health metrics

-Conducted research focusing on utilizing the opioid gene receptor to alleviate pain -Analyzed data and created charts for published research-Gained in-depth knowledge of pain research methodologies