Feb 2023 – Apr 2024 · 1 yr 2 mos · Remote

• At Appen, I work on developing AI content detection models that use behavioral and contextual features to detect LLM-generated content. My work was published in HCOMP 2023 as a Works-In-Progress paper.
• Additionally, I:
• developed internal benchmarking tools to estimate the performance of several third party LLM APIs for internal stakeholder deliberations
• write an internal newsletter reaching hundreds of views monthly on the state of AI and LLMs
• consult on LLM related projects/sales prospecting at Appen

Nov 2022 – May 2024 · 1 yr 6 mos · Remote · Remote

• In my free time, I create content and host events for Cohere on their Discord as an official Cohere Community Champion.
• I host livestreams, answer member questions, and create content related to using Cohere LLMs in projects.

Sep 2022 – Jan 2023 · 4 mos · Chicago, Illinois

• Got laid off from Appen, but I used this time to build out my portfolio and learn as much as I could about large language models.
• won first place at an NLP hackathon hosted by Cohere and Lablab.ai with my project Perfect Prompt
• served briefly as hackathon mentor and judge at lablab.ai
• contributed to an open source repository over at Hugging Face, some doc pages in Tensorflow for the Evaluate library
• became an official community representative for Cohere.ai, and wrote my first medium article for content creation
• gave a talk on NLP to some University of Nigeria students through pure serendipity!
• livestreamed a bunch of non-technical talks with lablab.ai on Twitch

Aug 2022 – Sep 2022 · 1 mo · Remote

• I work on the ML4Crowd team, where we aim to build models to solve problems related to monitoring the million-strong contributor crowd at Appen.
• Unfortunately, my role was affected by the Sept 2022 layoffs.

Oct 2020 – Jul 2023 · 2 yrs 9 mos · Park Ridge, Illinois, United States

• I assist students in becoming the best policy debaters they can be. This involves coaching them every week, attending tournaments to judge and prep, and researching articles on the current topic of protection of United States water resources.

Jul 2020 – Aug 2022 · 2 yrs 1 mo · Chicago, Illinois, United States

• I lead Speeko's data science efforts, ranging from user behavior analysis to speech recognition technologies. I work on refining Speeko metrics used in app to help users crush their public speaking goals, as well as general research and development into new data-oriented features.

Jun 2019 – Oct 2019 · 4 mos · Chicago

• My work at Speeko involves analyzing thousands of users within our application in order to inform and accelerate our growth. I use dashboards such as Mixpanel as well as powerful programming languages such as R and Python in order to understand our user base. I work on developing data visualizations and models of user behavior to convey my findings directly to our CEO and CPO, as well as to suggest routes for A/B testing changes to our app. Finally, I create models to predict user conversion and churn.

Oct 2019 – Jun 2020 · 8 mos

• Given two randomly selected research papers, is it possible to predict the probability that these are co-cited together in a future paper? I worked on answering this question at this research position, utilizing deep learning and natural language processing techniques.

May 2019 – Feb 2020 · 9 mos

• Going to college can be filled with uncertainty, and it is our responsibility as Peer Mentors to help mitigate these fears, and uplift incoming first year students with their transition. My responsibilities include serving as a resource through our online Wisr.io community, guiding students through monthly modules discussion responsibilities and expectations for college, and setting up times for discussion specific concerns with students if necessary.

Jan 2019 – Jun 2019 · 5 mos

• I tutored undergraduate students on a walk-in basis at the University of Chicago in introductory statistics courses. I find that being a tutor allows me to strengthen my knowledge base, as well as my ability to concisely explain concepts to students aloud. I found the most gratifying moments from this job where when I could help students completely grasp a concept they had trouble with at the start of my shift.

Sep 2017 – Jun 2018 · 9 mos

• As a Resident Dean Assistant, I helped the Resident Deans of Campus North East (a dormitory at U) plan, and execute events for 400+ students residing there. I made posters, managed a facebook group of about 700 members, pitched event ideas, and attended them in order to ensure they ran smoothly. We invited professors to give talks, showed movies, held study breaks, parties, and video game tournaments for students to relax.
• Responsibilities included advertising for events such as faculty speakers, study breaks, and themed parties, as well as staffing and engaging with attendants.

Jun 2018 – Aug 2018 · 2 mos · Greater Chicago Area

• In this role, I developed and validated speech analytics models for our mobile app, in order to facilitate providing automated feedback to users on their speech.
• I wrote scripts to analyze performance of our current approaches, I annotated data for our models, and I got a few projects of my own to work on. I was tasked with developing a new approach to conveying filled pause detection in our app. Starting with very little prior experience in NLP, speech signal processing, machine learning, and linguistics, I taught myself the basics of these skills and researched a new approach, generated datasets, and built a working model for filled pause detection, that became implemented in-app.
• I also assessed the quality of current data visualizations within the app, and suggested changes to optimize user interpretation of their own speech patterns. Finally, as we were apart of the 2018 Techstars Chicago cohort, I was able to attend entrepreneurship programming to expand and put into context my responsibilities in building out metrics for our app.

Jun 2017 – Sep 2017 · 3 mos · Greater Chicago Area

• I worked with Speeko to develop market assessments, contribute to speech metric analysis, and maintain outreach.
• During the internship, I taught myself Python and R and was moved to the metrics team to develop approaches at quantifying constructive speech feedback. This includes statistical analysis of audio features using R and data science packages, as well as generating basic machine learning models for interpreting and predicting public speaking performance. (speeko.co).

Jan 2015 – Jan 2018 · 3 yrs · Greater Chicago Area

• I judge policy debates at the high school varsity level, usually on weekends. This includes listening to speeches and providing constructive feedback to competitors, in order to better their own skills.