• Designed core IP and all AI/LLM models to build an automated GraphRAG growth marketing platform with a workbench that allowed users to create complex, no-code LLM-powered functions and agents. IP lowered CAC and raised conversion by synchronizing content at all layers of customer acquisition funnel.• Developed data engineering back-end on AWS, primarily through AWS Serverless (Lambda), DynamoDB, and AWS EC2 resulting in API endpoints for React/Next.js front end features including user management.• Built custom data science functions to create knowledge graphs from websites and documents, execute chained RAG queries and topic clustering, and automate website publishing and chatbot creation.• Presented to clients, board, and investors. Managed a team of software engineers and data scientists to create MVP of the product and scale to the first 10 customers.

• Went from handling minor aspects of analyses as a Temporary Associate to fully managing teams and organizing the workflow, and writing major parts of the quantitative analysis on six cases, some with potential damages in excess of $1 billion, as a Permanent Associate.• Work consisted mainly of analyzing large datasets primarily from the financial and health care fields to create 20-40 analyses to form the basis of legal arguments to support an expert witness’s report. Said analyses were centered on estimating damages in various scenarios using statistical modeling and machine learning frameworks on the data, causal inference, and assessing the statistical properties of the data itself. • Getting the data into a usable state often required the use of regular expressions or OCR scripts. • This position also often required producing data visualizations and summarizing analyses to explain the work to non-technical audiences. Primary programming languages employed were Python, R, and Stata.