I have over 5 years of experience in data science, machine learning, and programming. Since November 2022, I have been working as a Data Analyst at LS Direct, where I perform data analytics and data engineering tasks for some of the company's largest clients, helping the company quintuple the original budgets in some cases. I have developed both internal and client-facing automations using Python, SQL, and GCP, saving our team hours of time weekly. Whenever possible, I deeply favor programmatic solutions that streamline daily processes.

Previously, I worked at Lake Mary High School as an AP Computer Science teacher for two years, restarting the school's AP Computer Science A program (object-oriented and algorithmic programming with Java) and being the founding advisor for LMHS AI Club, introducing advanced students to deep learning with PyTorch and Linux. Prior to that, I taught algebra to disadvantaged communities at Vanden High School.

Before that, I've completed an undergrad in math and three graduate programs, one in math and yes, two in economics (applied and mathematical), the latest from UC Davis. As a graduate student, I got to work on computer vision applications and to teach statistical methods, econometrics, and graduate economic theory. I have always been drawn to linguistic and computational aspects of the world, and in the past, I've learned numerous languages (being reasonably fluent at roughly 7 at one point). Since then I have shifted my passion to deep learning and computer languages, especially LLMs and Python.

When I'm not at work, I'm either working away at LLMs, reading arXiv papers, following along with the fastai community, or hiking.

Things I Do

These days I'm mostly interested in building useful stuff with code

• Code (Python/PyTorch/SQL/GCP)
• Create value ($)
• Read arXiv papers
• Come up with ideas across disciplines
• Work in teams
• Drink coffee and hike

Projects

A sample of my projects from over the years

arXiv LLMs Assistant

arXiv-llms-assistant: 'Public' version of a project submitted as part of an AI competition at LS Direct, demonstrating the use and evaluation of RAG with on-prem data as a way to capitalize on applications of gen AI to privacy-restricted domains.

Applied exllamav2 6.0bpw quantization to run Mixtral-8x7B on personal GPUs with virtually no loss in quality and evaluate RAG pipeline configurations to find the best performing one (evaluation score ~90%).

Built an assistant to study the LLM domain, compare out-of-repository papers with papers in a personal repo (Zotero) and recommend recent papers to read along with lists of question/answer pairs.

Orchestrated a multi-faceted evaluation of RAG setups, implementing feedback mechanisms that led to the selection of the most effective configuration; this process improved question relevance, elevating overall project outcomes.

Built a command line utility to ask questions about the contents of uploaded papers and generate question/answer sets based on these papers.

A paper I co-authored

Went deep into economic theory and mathematical economics in the past...

Detecting anomalies in a statewide housing market with alternative data

Unsupervised learning (TDA) project to detect housing market anomalies built with Zach Stone for Haystacks.ai

Project Bird

GCP ETL project to pull data from eBird API 2.0 and display a map of recent bird observations in a given area: a small side project to practice GCP skills and see more birds.

Schedule an Airflow job to obtain data from eBird API 2.0 and store in Cloud Storage.

Clean the data via an Apache Beam Dataflow job and push the results to BigQuery.

Use a Cloud function to trigger the Dataflow job in Step 2 on file upload to Cloud Storage bucket.

Build geographic visualization with BigQuery Geo Viz.

Inventory classes with pytest and pdoc

Inventory-classes-with-pytest-and-pdoc: Inventory application for a tech video channel featuring computer builds with inventory classes in object-oriented Python, 90%+ coverage tests with pytest, comprehensive API documentation with pdoc, and packaging inside Docker.

Housing price prediction

Supervised learning and deployment project developed for NYC Data Science Academy

Education

MA Economics, University of California at Davis

MS Mathematics, University of Central Florida

MS Economics, University of Central Florida

BS Mathematics, University of Central Florida

Contact Me

dpopovvelasco{at}gmail.com