About the Author
David Yerrington is a seasoned Data Science and AI expert with over two decades of experience in transforming complex business challenges into actionable, data-driven solutions. As the founder of Yerrington Consulting LLC, David has led diverse teams to develop and deploy cutting-edge machine learning models, including large language models (LLMs) and retrieval-augmented generation (RAG) systems, driving significant business value for clients ranging from startups to Fortune 500 companies.
David's passion for technology and innovation has been a constant throughout his career. He has a proven track record of translating vague business requirements into measurable outcomes, developing advanced NLP models, and creating big data pipelines. His strategic initiatives have optimized user engagement, reduced churn, and enhanced operational efficiency.
Beyond his consulting work, David has made substantial contributions to the education sector. As a Global Lead Data Science Instructor at General Assembly, he has developed comprehensive curricula and case studies that are now taught globally, shaping the next generation of data scientists.
David is also an accomplished author and speaker, sharing his insights and expertise through articles, publications, and presentations at industry conferences. His work continues to push the boundaries of what's possible with data science and AI, driving innovation and growth for businesses worldwide.
Articles
What You Don’t Learn in Data Science Bootcamps
An article I wrote describing the need for high educational standards in bootcamps and specific points for students to consider to study up on afterwards for better job outcome success.
Articles
Published in LLMs, NLP Customizing AI: How to Leverage Your Own Data with Large Language Models Like ChatGPT for Business Applications
As a consultant, I often address how businesses can leverage advanced large language models like ChatGPT, tailoring AI with their proprietary data to generate unique insights and solutions.
Article
Devin, the AI Coding Wunderkind: A Skeptic's Take
Cognition Labs claims to have created the world's first autonomous AI software engineer, but is Devin truly a game-changer or just more AI hype?
ArticleLinks
AT&T Data For Diplomas, 3rd Place
We analyzed environmental theory to derive hypotheses on graduation rates. Using advanced techniques, we identified that household stability and economic security are key factors in improving nationwide graduation rates, with weather also playing a significant role. Conversely, school spending and low food access showed minimal impact on improving graduation rates.
Analysis
The Informed Company, ISBN 1119748003
Executive editor for the print version of The Informed Company
Publication
AT&T Data For Diplomas, Video Presentation
Video presentation explaining that educated households were NOT strong predictors of whether you would graduate high school.
Video
Hands on Parallel Computing with Dask and Pandas
In this session, you will learn how to work, hands-on, with the Dask framework to build scalable transformations to support analytic applications.
Training
AI+ Practical Advanced Pandas
Video sample of my session explaining how to style Pandas DataFrames. The complete series explains the more advanced features of working with Pandas.
Training
ODSC West Keynote: From Jupyter to Dashboards
In this presentation I instruct on how to create bespoke data visualizations prototyped in Jupyter but served in standalone containers.
Training
Deploying K-Nearest Neighbors /w Flask and Heroku
One of my first ever presentations I gave before being invited to develop General Assembly's data science immersive program in a leadership role.
Training
Analysis Arena Hackathon Promo
Promotional video for an educational hackathon. I used a combination of DAW, AI, and motion graphics produced in Adobe suite to create this promotional video.
PromoODSC West, Keynote on Data Visualization
ODSC presentation info workshop / talk on building workflows with Jupyter and production web applications.
Training
Intro to Machine Learning
Presentation for online workshop on the basics of machine learning, and what it can and can’t do.
Training
Topic Modeling Top Hip Hop Artists of All-Time
Presentation chronicling my journey surveying subject matter experts, acquiring data, and modeling rap lyrics from a list of moderated artists.
Analysis
Project Salmon: Modeling Online Dating Interactions
How I collected, cleaned, and built a predictive model that could accurately predict who would message me back upon first interactions.
AnalysisHands on Parallel Computing with Dask and Pandas
Technical assets and notebooks for working with data at scale with Dask and Pandas in Python.
TrainingHands on Parallel Computing with Dask and Pandas
Technical assets and notebooks for working with data at scale with Dask and Pandas in Python.
GithubPractical Advanced Pandas Workshop
Workshop I put together for AI+ for working with the more advanced features in Pandas in Python.
GithubData Visualization Workshop for ODSC West
In this workshop, I guide people through a set of use cases that guides students through taking their visualizations in Jupyter and deploying them to a hosted application solution.
GithubData Visualization Workshop for ODSC West - Slides
Slides for my workshop presentation on going from working in Jupyter to production web applications.
Presentation
Dose of Data: Analysis Arena #001
I organized and ran an education al hackathon focused on data analysis and machine learning with 600 successful registrations.
EventDose if Data: Analysis Arena Github Repo
Technical assets for participants including judging rubric, discussion forums, and Github posting templates.
Github
How to: Data Science (The Hard Way)
A brief overview of my career, how I got started working with machine learning, and a few recommendations for newcomers.
Training
Introduction to Recommendation Engines
In this presentation, I talk about collaborative vs content based recommendations.
Training
Online Data Science Courses are a Scam!
Many data science course creators promise free classes or super cheap courses. I've taken some of them, and they're great for a quick overview, but you'll want to take them on your own time because the pace is usually too fast for newbies. Most courses will be in video format with no interaction from teachers. If you learn best by doing, there's little value here. You can't get much practice on your own without paying or learning at a school that has hands-on labs and experienced instructors. Even then, it might not be enough to compensate for lack of experience.
Promo