Mary-Elizabeth Voss

Non-Academic

University of Southampton, Soutmapton, UK - Student Researcher | July, '20- Nov, '20

Master’s Dissertation: Used Bilevel Optimization to address Malawi’s deforestation crisis by allocating wood to meet demand when supply is restricted, while ensuring villagers can maintain economic welfare

• Built a model to emulate participant behaviors and the effects of supply, demand, costs, penalties, etc on:
• community leaders’ allocation of wood for harvest
• villagers’ behaviors depending on fine for illegal harvesting and travel costs
• the influence of altruism on villagers with respect to others and the environment
• Used AMPL to run trials, test sensitivity of parameters and model scalability, then analyzed outcomes to form conjecture
• Presented findings in a 14,000+ word paper with tables and graphs to convey data and results, and acknowledged shortcomings of model, recommended next steps and suggestions
• Novelties include: using Stackelberg-Nash to represent community leaders and households, Nash equilibrium for villagers competing for wood resources, and bilevel optimization approach for forest conservation at a household level

Academic

University of Southampton, Soutmapton, UK - Dissertation Researcher | July, '20- Nov, '20

Master's Dissertation (14,000+ words) using Bilevel Optimization to address Malawi's deforestation crisis by allocating wood to meet demand when supply is restricted, while ensuring locals can maintain economic welfare

• completing this dissertation taught me how OR methods are under-utilized in benefiting the environment, clean energy and socio economic issues, and convinced me to pursue a PhD to further explore the possible applications of the field
• notable novelties of the Bilevel Optimization model include:
• a modeled Nash equilibrium framework amongst villagers competing for wood resources
• a bilevel optimization for forest conservation where forest usage is studied at a household level
• a new case study involving a Stackelberg-Nash game between community leaders and village households east of the Namizimu Forest Reserve
• ran trial changing parameters (such as supply and demand of wood, number of villages, penalty for illegal harvesting) using AMPL
• cited 70 sources, including:
• academic papers, books, government plans, and news articles to understand economic landscape, dependency on wood for income and energy, and what measures were put into place to combat deforestation
• economic and operational research papers related to calculating and quantifying welfare and altruism, optimization methods and forest management

Comparably, Santa Monica, CA - Consultant (Product Manager) | Nov, '20- Jan, '21

Led the new product development of Comparably's Diversity Report feature

• purpose is to create surveys and reports to help client companies showcase their diversity statistics, attract more diverse employees, and help achieve diversity and inclusion goals
• worked closely with CEO to define product based on high-level ideas and feedback directly from clients (such as cultural appropriateness of survey questions based on region)
• took feature through all the product development life cycles to a live product, including:
• scope planning (minimum viable product and future features) with CEO, CTO, and designer
• concept (wireframes and data point gathering), UI, and UX design (product flow) with designer
• defining clear technical requirements for engineers
• countless cycles of product QA testing (to find and fix bugs and incorporate additional user feedback)
• project management (with help of JIRA) to keep engineers and leadership on schedule (and informed of delays from excessive bugs)
• delivered and launched the feature before departure, with plans in place to develop new features (such as a customizable report for clients)

Comparably, Santa Monica, CA - Product/Project Manager) | May, '18- Sept, '19

Develop Product Features

• converted original ideas (as well as abstract concepts from CEO) into tangible features that were compatible with existing data infrastructure and aligned with strategic goals
• translated features into clear technical requirements for engineers and managed release to market

SEO (search engine optimization)

• independently researched SEO tactics, and with guidance from my CTO and CEO learned several ways to improve organic traffic
• implemented strategies to improve the number of organic visits to website
• all the incorporated changes multiplied organic visits by 217% in 5 months

Project Management

• managed engineering team to ensure work was completed on time and communicated progress to CEO
• used growing technical knowledge to understand if an engineer under- or over-scoped a project and credibly pushed back on their estimated time to complete feature
• resulted in 2-3 features per week being released on time

Comparably, Santa Monica, CA - Data Analyst | July, '16- April, '18

Data Management

• responsible for growing database of company directory as well as validating company ratings
• tasks included data ingestion (sourcing and cleaning data, and deciding on ingestion), verifying user data (e.g. by comparing if salary data provided seems believable based on related metadata), and regularly tested site to check all calculations were working properly to display correct company culture data
• resulted in publishing validated data of 100,000’s of companies and users over 22 months

Data Analysis

• decided on how to use raw data to tell a coherent story for users to compare different companies
• researched and tested (using a sample set of data) how to make data comparable by applying statistical methods (such as z-tests, creating grouped distribution models, and Bayesian theorem)
• resulted in a product that allowed users to quantitatively compare company cultures

Feature Testing

• made sure new features work properly (throughout different development stages) to get them to market bug-free
• involved understanding how features should work, testing them and reporting issues to engineers
• repeated process until feature was ready for market
• helped bring ~100 features to market over a year, and tested live features to check they worked properly