Debt Consultants Group
Downtown Miami, FL
Software Engineer & Product Manager
December 2024 - April 2025
- Processed 400K+ daily records through an AI-driven ETL pipeline, ensuring scalable, high-throughput data ingestion
- Improved model accuracy by 60% and boosted conversion rates by 25% through data-driven decision optimization
- Developed a real-time SMS analytics platform using Python, Flask, and PostgreSQL for instant performance insights
- Reduced client-lender ledger generation time by 98% (from 1 hour → 1 minute) through automation and workflow refactoring
UHealth | Sylvester Comprehensive Cancer Center
Miami, FL
Data Engineer
August 2023 - August 2024
- Architected and maintained a 10,000+ sample tumor database powering AI-driven diagnostic and computer-vision models
- Ingested and transformed 20+ TB of histopathology imaging data across Olympus and Aperio formats for large-scale analysis
- Supported 300+ clinicians and researchers via cloud-based digital pathology systems, improving accessibility and workflow uptime
- Enhanced data reliability and research collaboration through secure, HIPAA-aligned infrastructure and governance protocols
UHealth | Epic Systems
Remote
Software Instructor
January 2023 - June 2023
- Trained 60+ research coordinators and clinicians across modules, driving cross-functional proficiency
- Achieved 95% user adoption through 10+ live Microsoft Teams training sessions and follow-up support
- Reduced staff onboarding time by 25% through workflow streamlining and tailored process documentation
- Partnered with Epic module teams to streamline and align clinical trial workflows
National Institute of Health
North Druid Hills, GA
Neuroimaging Analyst
January 2019 - June 2021
- Optimized predictive ML models (scikit-learn) on longitudinal fMRI data from 40+ stroke patients, improving classification accuracy by ~15% in modeling motor network recovery
- Developed custom UNIX shell scripts (awk, grep, sed) to parse and structure behavioral log data, automating feature extraction and reducing preprocessing time by 30%
- Labeled signal components and trained logistic regression classifiers to separate neural activity from motion, cardiac, and physiological noise, enhancing BOLD fMRI data reliability
- Automated unsupervised preprocessing workflows for fMRI time-series data, improving reproducibility and end-to-end pipeline efficiency
Emory University School of Medicine
Atlanta, GA
Data Engineer
January 2018 - January 2019
- Processed 250GB+ multimodal neuroimaging datasets through streamlined ETL workflows, enhancing throughput and standardization
- Reduced data processing time by 30% through workflow optimization and batch automation
- Converted DICOM to NIfTI formats using Bash and FSL utilities, ensuring reproducible voxel dimensions and metadata consistency for fMRI research
- Prepared TMS and fMRI datasets for motor cortex excitability and cognitive-decline analyses, improving data quality for downstream modeling