I'm currently working to build Coco Alemana – an IDE for Data Science.https://cocoalemana.com

During my tenure at CultureX I built and oversaw several projects. My main duties revolved around the development of software for both internal and external use.Here is an overview of a few of the projects:- Built a core services API in Python + AWS Lambda, and an associated Electron desktop app to allow for internal processing of data in a user-friendly way.- Modified a core classification algorithm by converting to C++ and reduced memory usage by 70% and runtimes by 90% over existing Python version.- Designed and built a suite of data processing tools to handle small and large datasets (100m+ records) using Apache Spark (Python) on Amazon EMR Serverless.- Oversaw the development of several ML models, including a high performance XGBoost model which is able to train, and evaluate 200+ models in parallel, while taking less than 60 seconds to produce results (with cold start). (AWS Lambda + Step Functions)- Designed and developed a default-deny security policy evaluator based on AWS IAM syntax.- Researched, architected and oversaw the development of a system to perform just-in-time (JIT) analytical queries for <1s response times based on user-provided queries.- Managed the security posture of all infrastructure, including automated logging, zero-trust policies, and automated infrastructure security tests via AWS CDK.- Developed and maintained a set of AWS CloudFormation stacks for all pieces of infrastructure, allowing for easy deployment and replication of entire infrastructure components in other accounts.- Developed and oversaw continued development of a SPA React (with Typescript) web app which allowed our clients to access and make queries against processed versions of their provided data.- Oversaw the development and deployment of a clause-accurate sentiment model. Sourced labelers, and coordinated with ML engineer to develop and evaluate model accuracy.Please reach out for more information.

Worked along-side Dr. Donald Sull to measure corporate culture in a quantitative fashion. Built big data pipelines and worked with the MIT Sloan Engaging HPC cluster to process millions of free text Glassdoor reviews. Used those reviews for the Culture 500, an MIT Sloan Management Review project focused on comparing roughly 500 of the United States' biggest companies on various dimensions of culture. Performed various analyses to ensure the integrity of our data before publishing. Performed various correlation and "fit" analyses to understand the grouping of our NLP topics.

Build a side project travel app. PHP, CSS and HTML.