Advanced analytics, natural language processing, and computer vision to automate InsurTech workflows.Automated Document Processing Solutions: - Built computer vision algorithm (convolutional neural network and autoencoder) to identify all document types found within closing packages with 99.7% accuracy in collaboration with senior staff data scientist. - Upgraded NLP based machine learning system (BERT, transformers) to classify fees found in closing disclosures in support of business expansion from mortgage refinance to home buying.- Deployed bespoke RegEx based document parsing solution to support custom workflow for top referral partner. - Owned data curation model using NLP and logistic regression to create updated training dataset that lead to a tenfold improvement in document identification rate in production. Business Intelligence, Forecasting and Analytics:- Developed Monte Carlo simulation to forecast the impact of existing and future document processing solutions on overall automation rates and delivered findings to C-suite and upper management to inform product roadmap. - Administered Looker dashboards monitoring data science product level performance and automated components of the tech stack to improve operational efficiency and reduce costs by lowering the number of long running queries from over 40% to less than 2%.- Designed A/B test to measure the impact of automated document processing on overall home closing times.- Mentored summer intern and supervised business intelligence training

Engaged in rigorous and immersive project-based training in data mining, data visualization, and machine learning at Metis, an accredited 12-week full-time data science program.Projects:# 1. Canada’s Response to Economic Impacts of COVID-19Deep learning & natural language processing to identify clusters of Twitter conversations about Canadian government financial aid programs during the pandemic. Tools: Transfer learning from Godin 2019 Twitter embeddings library, FastText to create document embeddings, K-Means and HDBSCAN to identify clusters, VADER for sentiment analysis, EDA and visualizations using Matplotlib, Seaborn and Yellowbrick. # 2. Actionable Insights from Lululemon Product ReviewsNatural language processing & unsupervised learning exploration of customer reviews of lululemon’s best-selling sports bras. Created a streamlit app to curate reviews for customer product needs.Tools: topic modeling with LDA, LSA, NMF; visualization of topic similarity with t-SNE, NLP preprocessing (TFIDF, CV, Spacy), web scraping with Selenium, Streamlit. # 3. Deadly Police EncountersBuilt a classification model to identify where deadly police encounters are likely to happen based on socio-economic characteristics of the communities in which they occur and created an interactive map of the United States to visualize their geographic spread.Tools: Classification Algorithms (Logistic Reg, KNN, Random Forest, Naïve Bayes, Support Vector Classifier), ROC-AUC Curves, Sci-kit Learn, Tableau.# 4. Chinese Box Office SuccessUsed linear regression to determine features of films that lead to commercial success in the Chinese market and applied model to all films released in China from 2016-2019 to predict what films were most likely to have highest revenue.Tools: Linear Regression, Webscraping with Beautiful Soup, Python, Pandas, Statsmodels OLS, LassoCV, Matplotlib, Seaborn.

- Leverage superior knowledge of Chinese diplomatic history and Mandarin language skills to provide insight into China’s role in shaping the modern international order.-Collaborate with team of 12 international experts to identify and investigate taken-for-granted concepts in international politics to write a new conceptual history of international relations.-Team awarded $1.1 million USD by the Research Council of Norway.

- Promote knowledge sharing about China’s global ambitions and their implications for Swedish foreign policy across academia, government, civil society and business sectors - Participated in panel discussions on China’s Global Role in the Covid-19 Crisis and the Belt and Road Initiative for audiences of 150-200 diplomats, academics, businesspeople and journalists - Interviewed on film providing analysis of China’s intentions to change the international order for the Global China Interview Series

- Manage multiple lines of independently designed social science research projects on China’s foreign policy, Tibet, the South China Sea and systemic global change - Collaborated with 3 senior level colleagues to secure $1.8 million USD from the Bank of Sweden Tercentenary Foundation- Authored 6 original articles averaging 8,000 words each, 3 published as working papers- Presented research findings at 4 invited lectures, 5 invited workshops and 5 international conferences - Co-organized and planned 60-person international conference at Lund in June 2020 (postponed due to COVID-19)- Awarded 2018 prize for Best Paper in Historical International Relations presented by a Graduate Student at the 2017 International Studies Association Annual Meeting (most respected and widely known scholarly association in International Relations)

- Managed team of 3 co-teachers, designed and taught graduate-level course of 14 students - Co-taught undergraduate course of 15 students - Prepared lectures, managed student discussions, provided students with progress reports and extensive written feedback - Advised late-stage PhD student and evaluated 200-page doctoral dissertation manuscript

- Employed training in research design and methodology to conduct state-of-the field analyses, identify avenues of research and independently design new studies to address them - Located, digitized, systematically managed and analyzed thousands of Chinese and English language diplomatic documents collected from sources in China (PRC), India, Taiwan (ROC) and the UK during 15 months of fieldwork - Synthesized data to produce reports including 200-page PhD dissertation and presented research findings at 10 international meetings and 1 invited public lecture - Published peer-reviewed article at leading journal on contemporary Chinese affairs cited by at least 13 other peer-reviewed publications - Led organization of 5 panels at international conferences- Awarded grants including US Department of Education Foreign Language and Area Studies Fellowship ($22,000 USD); Einaudi International Research Travel Grant ($4,600); Lam Family Award for Research in South China ($4,500); Cornell Graduate School Research Travel Grant ($2,000); Lee Teng Hui Fellowship in World Affairs ($12,500); C.V. Starr Fellowship in East Asian Studies ($12,500)

- Assisted faculty members in teaching 6 Cornell undergraduate courses ranging in size from 18-200 students with classroom instruction, exam creation, grading and record keeping- Prepared lectures, organized and led discussion sections for groups of 6-18 students - Received excellent student evaluation scores averaging 4.41/5 - Served as editor of student run peer-reviewed journal Cornell International Affairs Review

- Cultivated donor relations and business partnerships - Managed logistics of the organization’s largest fundraising event and met fundraising goals of $250,000 despite the global financial crisis - Edited online weekly journal and contributed to brand awareness campaigns in traditional and social media outlets