Professional Background

I began my career as a software engineer working on testing automation software in C, Java, and Python. Later, as a consultant, I built backend applications using Python, Golang, PostgreSQL, MongoDB, and Google Cloud products.

Founder, Editor-in-Chief

Since LearnDataSci's inception almost a decade ago, I've worked with many talented writers and editors from varied technical backgrounds to help them create exceptional data science content.

I'm passionate about the science of learning, and I focus on implementing effective learning strategies and memory techniques in LDS's articles. My writing and editing emphasize clarity, step-by-step thought processes, and practical examples to make complex concepts accessible.

Course Reviewer

In addition to directing the content strategy, I evaluate roughly 50 new data science, ML, and AI courses each year. Since 2015, I've reviewed over 500 courses, dozens of books, and several bootcamps to help readers navigate the online learning landscape.

Based on my recommendations, tens of thousands of readers have enrolled in courses through LearnDataSci.

Expertise

Most of my expertise is in software and the engineering side of data science, ML, and AI, where I focus on Natural Language Processing, data engineering, and information retrieval.

Education

I hold a Bachelor's degree in Mathematics from Framingham State University, with a minor in Computer Science and Chinese. As I've enrolled in many online courses I've reviewed for LDS, I've accrued many skills in data science, ML, and AI.

Projects

I'm usually working on a software or data science side project. Here are a few past and present examples:

Crypto Price Prediction Software

I utilized Twitter and Reddit data collected via Python APIs (before the API changes) and stored in Postgres on AWS. I implemented custom social media language tokenization, lemmatization, and spam/bot filtering on the training data and fed it to various models. The best predictive performance came from ensemble models, validated with a train/test split.

Cardinal App

Cardinal is a multi-language dictionary and language learning app. I built a Golang backend to retrieve any of the 20 million word definitions and usages in under 10ms. To achieve this efficiency, I minimized roundtrips with stored procedures in Postgres and highly optimized search queries. I implemented a custom text similarity score in PL/pgSQL, but I have plans for future improvements using newer retrieval techniques.

Course Aggregator

LearnDataSci is supported by software I built with Python and serverless Google Cloud products (Functions, Tasks, Storage). Thousands of courses and programs are collected and stored in raw form, transformed and normalized, and then loaded into Cloud SQL. The public-facing app, under construction, will help learners build curriculums, find suitable courses, and connect with others.