Currently an Extern @Breaking Games·🟢 Open to work
I am focused on building my expertise in Artificial Intelligence and ready to tackle new challenges in tech.
Explore my projects focused on Artificial Intelligence and Database Architecture, showcasing my developing skills and hands-on experience.
Architected a “Digital Twin” chatbot using a RAG (Retrieval Augmented Generation) pipeline to emulate my persona based off my general profile, experience, and education with high accuracy. Implemented a persistent vector database using ChromaDB and use OpenAI embeddings model to enable semantic sear
Built an AI agent workflow with Lindy to automate the collection of air quality data from the EPA website using their API. Agent can be triggered adhoc or follow a set schedule with a human-in-the-loop guardrail to confirm and authorize the data collection process. When the task is done, the agent e
I am focused on building my expertise in Artificial Intelligence and ready to tackle new challenges in tech.
I am an emerging professional specializing in Artificial Intelligence, currently exploring opportunities to enhance my career. With a keen interest in SQL and Database Architecture, I'm eager to develop my skills and make a competitive mark in the tech industry.
Externships
SQL & Database Architecture Externship with Breaking Games
Breaking Games
Experience
Technology Architect Analyst
Accenture · Jun 2022 - Dec 2024
Education
University of Puerto Rico - Mayaguez
BSc. Chemical Engineering · Class of 2013
University of Rochester
MSc. Chemical Engineering · Class of 2015
Thinkful
Online Bootcamp Data Analytics · Class of 2021
Skills
Architected a “Digital Twin” chatbot using a RAG (Retrieval Augmented Generation) pipeline to emulate my persona based off my general profile, experience, and education with high accuracy. Implemented a persistent vector database using ChromaDB and use OpenAI embeddings model to enable semantic sear
Built an AI agent workflow with Lindy to automate the collection of air quality data from the EPA website using their API. Agent can be triggered adhoc or follow a set schedule with a human-in-the-loop guardrail to confirm and authorize the data collection process. When the task is done, the agent e
The pet medication scheduler is an application built with AI assistance that lets pet owners set, manage, and visualize the dosage calendar for medications. Features include input of pet information, setting medication details like dose and duration, clearing or updating medication data, downloading
The pet medication scheduler is an application built with AI assistance that lets pet owners set, manage, and visualize the dosage calendar for medications. Features include input of pet information, setting medication details like dose and duration, clearing or updating medication data, downloading a PDF version, and providing visual feedback to let the user know of any errors.
- Delivered a data-driven store expansion recommendation by normalizing store performance for tenure (revenue per active store-month) and modeling city-level market saturation, identifying an under-served, high-performing city and presenting the finding with supporting methodology. - Built an end-to-end sales analytics pipeline in Python (pandas) across 5 relational data sources — 829K+ transactio
✅ Verified by Extern · ⏱️ In progress
Turn six messy CSVs into an analytics database and a Q4 dashboard that drives real business decisions. SQL + Claude.
The externship ingested six messy CSV sources and converted them into an analytics-ready SQLite database. The work connected to the provided database, used Claude to craft and refine SQL prompts, and ran queries across six tables to answer the module's sample business questions. For the schema work, a dim_product table and consistent product slugs were created and joined to sales data to produce
I connected to the supplied SQLite database, used Claude to refine SQL prompts and queries across six tables, and designed a dim_product table with slugs to unify four messy spreadsheets into a single product key linked to sales data.
On the Breaking Games externship I connected to the provided SQLite database, used Claude to help craft and refine SQL prompts, ran queries against the six tables, and produced query results that answered the module's sample business questions.
A messy set of four spreadsheets was unified by designing dim_product, creating slugs for consistent product IDs, and joining the dimension to sales data. The work produced a single product key and connected tables for downstream queries.
I analysed raw sales, referrals, and Meta campaign data, designed a channel taxonomy, ran SQL queries for key metrics (top products, total spend, checkout sessions), and produced a Q4 ad allocation plan with campaign-level ROI and recommendations.