Leonel Marquez

Responsible for the data area of a new consulting branch (client: public water utility, ~580 employees). • Built the complaints & requests KPI system from scratch: 2,000–3,000 records/month, presented monthly to management. • Built and deployed on my own a multi-client Dash app on AWS Lightsail + Supabase (PostgreSQL) for reputation monitoring, in production and used directly by stakeholders. • Built an offline field-survey system (Power Apps, Power Automate, SharePoint) for ~30,700 records.

Implemented the company's first KPI reporting system via Meta API, expanding historical data from 3 months to nearly 3 years. Designed Power BI dashboards (DAX, Power Query) with business KPIs and Instagram metrics, incl. competitor benchmarking. Built a Python/Streamlit app for real-time order tracking, replacing a paper-based process. Developed an inventory module with automatic USD/COP/VES conversion. Presented insights to management and non-technical stakeholders.

Delivered online group classes (up to 6 students) over 3.5 years, maintaining a consistent 4.8–4.9/5 student satisfaction score.

Delivered 5,096 one-to-one lessons to 309 students with diverse learning goals, incl. DELE/SIELE preparation and Business Spanish, achieving a 4.9/5 rating from 54 verified reviews.

Final project: brain MRI image classification using PCA and SVM. Evaluated 8 machine learning models on a dataset of 5,712 images across 4 categories, achieving 94% accuracy (F1-score: 0.94), after reducing dimensionality from 40,000 to 530 features.

Thesis: analysis of accounting processes at a manufacturing company (Tenería Rubio C.A.) and design of an Excel-based solution to streamline purchase order tracking through inventory control and automated reporting.

Thesis: developed a temperature control system using Peltier cells with remote TCP/IP management via Arduino, Ethernet, and a PHP/MySQL web app. Operated across a 38°C range between cold and hot phases, validated using Wireshark network traffic analysis.