Jobeth Muncy

Washington, DC · (719) 684-5660 · jobethmuncy@gmail.com

Mountain girl at heart, always seeking a new adventure. A career working with people has given me an in-depth human intuition that allows me to adapt and understand new environments. I thrive on finding new ways to bring out the best skills in those around me. A natural curiosity and a logical mindset enables me to solve complex data challenges with a thoughtful approach.

* currently in Washington, DC but willing to relocate to Portland (OR), Denver, and Seattle

Experience

Data Scientist | Data Engineer

General Assembly
  • Created a classification model to sort text from Reddit comments to determine which category they fit into: wine or cocktail. Used an API and Beautiful Soup to compile 100K comments for two different subreddit discussions. Applied Natural Language Processing (NLP) techniques to remove 43% of the original content leaving only meaningful words. Trained and tested a Naive Bayes model that correctly classified 87% of comment segments.
  • Collaborated FEMA and New Light Technologies to take over 63K scientific articles about coronaviruses and created an application that allowed the user to find articles for their research with a direct link to the paper. Used the Doc2Vec and Latent Dirichlet Allocation to sort the articles by topic and key words.
  • Created a wine recommender allowing the user in input any words about flavor, region, or style and returning similar items. Wines were scraped from four different retail sites using Python and Beautiful Soup. The FastText library within Gensim took user input and generate similar words to the input string and return recommendations using a Flask app with price, country, and direct link attached to the name.
March 2020 - June 2020

Bartender/Server

Tail Up Goat
  • Utilized communication and deduction to guide and enhance dining experiences at a Michelin starred restaurant. Executed flawless service adding personal and thoughtful touches throughout the meal resulting in an increased retention of repeat customers.
  • Curated unique dining experiences for guests with custom beverage recommendations based on expert knowledge of cocktails, beers and wine.
  • Responded to late or unannounced guest arrivals, dietary restrictions or aversions, kitchen and bar delays with an immediate action plan. Pivoted the situation to be unnoticed by the guest and creating a positive outcome leaving them with an uninterrupted, magical experience.
August 2016 - March 2020

Education

General Assembly

Data Science Immersive - Washington DC
Completed a 500+ hour immersive program developing skills in Python, Scikit-learn, SQL, NLP with extensive use of the Gensim library, webscraping, data cleaning and analysis, Machine Learning, Scala, Spark, and Neural Networks. Emphasised both statistical analysis and client facing interpretability of data findings.
March 2020 - June 2020

Skills

Programming Languages & Tools
  • Python
  • Pandas
  • Numpy
  • SQL
  • Spark
  • Scala
  • Data Visualization
  • Machine Learning
  • Natural Language Processing (NLP)
  • Gensim
  • Webscraping
  • HTML

Projects

Pick me a Wine!


The world of wine can be confusing to navigate. Every country has different rules and regulations for labeling and selling wines. A wine label itself can be difficult to read. If someone knows one or two words about a flavor profile, producer, grape, vineyard site, region, appellation how do they find that wine or other similar ones? I created a wine recommender that will take text and recommend wines that are available to purchase with a link to the site so they can buy and explore.
I scraped 12,000 wines from Astor Wines, Chambers Street, Twenty Twenty Wines, and Wine.com to give a broad selection of wine for the user to pick from. Gensim FastText was utilized to take user input and find words that are the most similiar to the segment and return wines that match the description. Direct links to selected wine is provided and each site has their own "similar wines" function to continue expanding user selection.
To see this project, click here

Indexing and Outbreak


COVID-19 brought the world to a stand still. To aid in the race for a vaccine and better understanding of the virus, the Semantic Scholar team at the Allen Institute for AI created CORD-19, the COVID Open Research Dataset, in partnership with Georgetown University, Microsoft Research, Amazon Web Services, the Chan Zuckerberg Initiative, the National Institutes of Health, and the White House created a database of over 63,000 scientific papers regarding coronaviruses and other similar outbreaks from 1957 to today.
How would a researcher find what they are looking for quickly with so much data? In collaberation with FEMA, Clay Carson, and Cynthia Chiang, we utilized both Doc2Vec and LDA within the Gensim library to search by keyword or topic. This is intened to expidite the research process in the race for a vaccine.
To see this project, click here

What are they talking about? Cocktails or Wine?


I collected 100,000 comments using webscraping and the reddit API from both the cocktail and wine subreddit categories. The text was cleaned by removing HTML, non-letters, all words lowercased and tokenized, stopwords removed, and lemmatized. Then all words were turned into numerical values using CountVectorizer and TFIDF. Several classification models were trained to best classify unseen text into the appropriate subreddit category. At the end of the day, the combination of CountVectorizer and Naive Bayes Classifer was the best combination at 86% accuracy. As a fun experiment on the model, I had friends submit text to me about wine or cocktails and the model correctly classified 14 of 15 lines of text.
To see this project, click here

Interests

A Colorado native, my passions are exploring the wilderness, hiking to new mountain tops, finding remote alpine lakes, and chasing marmots at high elevations.

From my past career in restaurants, I love experiencing the creativity of chefs, vintners, and bartenders. I spend vacations seeking out historical bars or favorite vineyards. I collect cookbooks from my favorite meals for at home attemts at their magical bites.