Jennifer Brookhart Email & Phone Number

Salt Lake City, UT, US

Purchase, NY, US

Purchase, NY, US

Purchase, NY, US

Austin, Texas, US

Provided data insights to third-party news outlets to support their stories and raise JungleScout brand awareness. Also supported internal teams in publishing information for sellers. Here are some of my favorite pieces:1) NYT's "He Has 17,7000 Bottles of Hand Sanitizer and Nowhere to Sell Them"2) WSJ's "Amazon Scooped Up Data From Its Own Sellers to.

Austin, Texas, US

• Orchestrated collaboration between data science and product teams, leading to new tools in product forecasting, competitor identification, and click and conversion rates, which help third part Amazon sellers manage.
• Liaisoned between data science and marketing to provide analysis for thought leadership content, including the impact of Prime Day on sales, and global shifts in suppliers during US-China trade war. Work picked up by.
• Headed team to estimate search volume on Amazon for 20M+ keywords, adopting systems to AWS framework with Apache Spark techniques to process 2.5TB of data. Final stacked model generates on-going estimates with 95%+.
• Optimized a product discovery tool with a Naive Bayes classifier and threshold cutoffs, resulting in a 60% reduction of the false positive rate.
• Improved the accuracy of core sales estimator model by 20% in the first four weeks of employment. Accomplished this by expanding the data pool to avoid selection bias and switching from calibration curves to a machine.

San Francisco, CA, US

• Created Landscope end-to-end, which geocodes plant growing information to users’ local addresses to return list of best plants for their area, with resulting accuracy 3 orders of magnitude higher than similar web apps.
• Wrote advanced SQL queries combining 7 datasets from APIs and 21.6 GB of web-scraped plant data for collaborative filtering in Landscope.
• Designed and led tutorials on statistics and on scrapy, a python library for collecting data from unstructured sources.
• Consulted on 14 NLP and supervised machine learning projects.

• Managed 5 person team to establish the journal. Oversaw first 2 issues.
• Designed and maintained journal websites, drafted mission statement and protocols, and edited manuscripts.
• Acted as liasion between authors, editor, and press

Madison, WI, US

• Team Management:
• Directed a team of 11 researchers in analyzing 200,000 congressional documents, using a semi-supervised classifier and Bayesian hierarchical topic model, which designated the level of policy or district focus of each.
• Oversaw 16 undergraduate and 3 graduate research assistants on data collection and analysis.Statistical Analysis:
• By optimizing python code for sparse matrix calculations, improved discipline-standard topic model to be fully non-parametric and run 20x faster.
• Crafted a new dynamic topic model using python and C++ to analyze the interactions between multiple corpora, including 2.6B tweets, 1.1M news stories, and 59K congressional documents.
• Parallelized analyses to run over multi-core CPUs and the Wisconsin computing cluster.Data Wrangling:

Madison, WI, US

• Independently designed and taught 6 courses, ranging in size from 12 to 140 students.
• Secured university funding for a lab-based blended learning seminar.
• As head teaching assistant, assigned discussion sections and organized exam schedules and grading.

• Organized 37 workshops, including travel, schedules, and event promotion.

• Coordinated program activities for the start-up, including public events, courses, community education, and research projects.
• Helped to permanently establish the Forum with a $2M endowment.
• Won fellowship to travel to Pasadena, CA, for two-week seminar
• Taught statistical analysis and research design in two honors seminars

Bachelor Of Arts - Ba, Political Science And Government

Doctor Of Philosophy - Phd, Political Science And Government