TECHNICAL SKILLS:
Primary Languages: Python, SQL
Secondary Languages: R, Java, Tableau
Data Engineering: HDFS, Spark, Neo4J, MongoDB, JSON, API, Web Scrape
Python Packages: Sklearn, Scikit - Learn, Numpy, Pandas, Tensorflow, Matplotlib, Plotly, Seaborn
Machine Learning: K-Nearest Neighbors, Support Vector Machine, Principal Component Analysis, Density-Based Scan, Latent Dirichlet Allocation
WORK EXPERIENCE:
Confidential, AI
Data Scientist
Responsibilities:
- Developed architecture for dynamic anomaly detection and clustering in large-scale, streaming graphs
- Developed machine-learning conversational guide based on utility of conversation
- Including features and models drawn from domain knowledge in psychotherapy, sociology, and neuropsychology
- Processed lyrics and generated rhyming text seeds with Neo4J and Python
- Developed Triphone2Vec rhyme and accent detection algorithm
Confidential
Data Science Interns
Responsibilities:
- Identified new market coverage possibilities for venture lending and methods for machine learning to recommend likely deals of interest to SVB staff
- Graph-based analysis of venture capital ‘social networks’
- Developed recommender system capable of predicting early stage deals of interest to critical late stage investors
- Performed random walk clustering of investor networks and validated the results via known networks
- Trained ERGM (exponential random graphical model) as a means of investigating investor closeness
- Utilized investor participation data and neural network to predict likelihood of negative venture debt performance as a classification problem
- Utilized matrix decomposition for broad, sparse dataset
- Training accuracies averaged 80%
- Optimized sample portfolio performance via elimination of high-risk deals based on likelihood estimate and business decision-making utility
- Test portfolio performance improved by over 10%
- Developed query-ready graph prototype, including integrating internal and external data sources
- Utilized Density Based-Scan to identify networks of high-value investors
Confidential
Senior Financial Data Analyst/Data Operations Specialist
Responsibilities:
- Sole analyst and data ops specialist for Confidential -mortgage mortgage backed securities and applications. Responsibilities included:
- Validating new cashflow models against origination documents. Importing new model and data into Structured Finance Workstation deal library and SQL database. Maintaining relational database.
- Providing quality assurance for monthly releases of data from the point of trustee release of data through SQL transformation to deal library and final published approval for client use
- Client-facing communication, including providing domain expertise and customer service support
- Extensive data munging in SQL and Excel
- Trained and audited new hires and outsourcing team
- Developing QA procedures and documentation for monthly data
- Developed, wrote, and implemented QA logic, successfully automating 70% of approvals from a previously manual process
- Oversaw 100% growth in the size of the Confidential library of deals
- Oversaw universal data backfill of Confidential historical data
- Designed and implemented projects to supplement and improve upon trustee property level data. Included extension of property default logic and development of key fields such as net-operating income and loan-to-value ratios
Confidential
Research Assistant
Responsibilities:
- Researched, compiled, formatted, and summarized data sets based on potential for use in economics textbook, selecting for a combination of student interest and relevance to econometric study
Confidential
Assurance and Advisory Intern
Responsibilities:
- Audited Charles Schwab website transition, trading process, and authorization, including trading floor process walkthrough, for Financial Services Risk Management and Internal Audit team
