← Back to contact Paul Ramirez-Lopez
Computer Vision Engineer
Applied Computer Vision Engineer with approximately 3.5 years of paid R&D experience developing practical computer-vision and image-processing workflows for PCB and microelectronics inspection, hardware assurance, and automated inspection systems. Experience spans object detection, OCR and text analysis, component classification, anomaly-detection research, dataset engineering, large-scale annotation, multimodal imaging research, customer/SBIR delivery, and technical project leadership.
Based in Florida; Open to Remote and Relocation.
U.S. Citizen; Clearance Eligible.
Selected Experience
Oct 2022 - Oct 2024
Gainesville, FL
Caspia Technologies
Research & Development Engineer
Applied CV/R&D role focused on PCB and microelectronics inspection, AutoBOM and counterfeit-detection research, dataset engineering, and technical delivery.
- -Contributed to AutoBOM and counterfeit-detection R&D programs that converted PCB imagery into structured component, marking, text, orientation, and inspection metadata through computer vision, OCR, classification, and visual-association workflows.
- -Developed and supported PCB inspection subsystems including YOLO-based integrated-circuit detection, surface-mount component classification, board/component-marking recognition, OCR/orientation workflows, and component-to-reference-designator association.
- -Expanded a PCB vendor-logo and marking dataset to ~7,000 instances across ~1,300 marking types and contributed to ~300,000 total manual annotations spanning components, markings, text, orientation, reference-designator relationships, and OCR-support metadata.
- -Investigated counterfeit and defect-detection approaches under severe data scarcity, including synthetic-data augmentation experiments and image-based surface inspection workflows for microelectronic components.
- -Authored an independent ~40-page multimodal inspection feasibility study covering SWIR, infrared, X-ray, and other sensing methods; helped justify acquisition of a high-end SWIR imaging system and established initial experimental PCB imaging workflows.
- -Produced weekly technical presentations and monthly SBIR-style deliverables documenting model development, dataset status, technical risks, and project milestones; onboarded and coordinated junior engineers across Physical Assurance R&D workstreams.
- -Served as NSF I-Corps Entrepreneurial Lead for a seven-week commercialization program, conducting 100+ stakeholder interviews across electronics manufacturing, inspection, procurement, recycling, and supply-chain security.
May 2021 - Oct 2022
Gainesville, FL
Florida Institute for Cybersecurity Research
Research Assistant
Hardware-assurance research foundation in PCB inspection, image processing, dataset construction, and ML experimentation.
- -Supported early hardware-assurance and PCB inspection research that later expanded into AutoBOM and counterfeit-detection workstreams, contributing across dataset construction, annotation, image processing, and machine-learning experimentation.
- -Developed PCB image auto-scaling and resolution-estimation workflows using reference-object detection, PCB boundary detection, edge/contour analysis, and known physical-reference geometry to estimate image PPI under inconsistent capture conditions.
- -Built the initial PCB vendor-logo and marking dataset, curating ~3,000 examples across ~800 marking types for downstream vendor-logo, standards-marking, and board-marking recognition.
- -Produced ~70,000 PCB annotations across components, markings, and text-related metadata, including localization, orientation, and class labels.
- -Explored GAN-based synthetic counterfeit/defect generation for hardware-assurance datasets, identifying scarce confirmed negative examples and high visual variability as primary limitations to usable augmentation.
Jul 2018 - Aug 2021
Gainesville, FL
Admiral
Account Development Representative → Customer Success Representative
Earlier SaaS customer-facing experience in outbound sales, customer success, issue resolution, and account communication.
- -Supported outbound sales, customer success, issue resolution, and account communication for a SaaS platform serving digital publishers.
- -Built early experience translating customer needs and product issues between users and internal teams.
Featured Project
Maritime Vessel Detection
An audit-driven remote-sensing detection workflow built around DOTA v2.0, xView, and HRSC2016-MS. The work emphasizes annotation quality, parent-space coordinate handling, OBB conversion, tiling, curation, controlled augmentation, and multi-seed evaluation.
Technical Areas
Computer Vision Tasks
Object Detection · Image Processing · OCR and Scene Text Detection · Feature Extraction & Association · Anomaly Detection · Image Segmentation · Training & Inference Optimization
Models and Frameworks
PyTorch · OpenCV · Ultralytics YOLO · SAM-Assisted Labeling · Tesseract, EAST, and CRAFT
Data and Experimentation
Dataset Curation · Annotation QA · Oriented Bounding Boxes · Segmentation Masks · Tiling Pipelines · Experiment Design / Ablation Testing · Domain-shift Analysis
Deployment and Engineering
Python · Docker · FastAP · SvelteKit · ONNX Runtime · Git · Linux
Education and Training
B.S. Business Administration, Management; Minor in Entrepreneurship
University of Florida · Gainesville, FL · May 2020
- Computer Vision Specialization - University of Colorado Boulder
- PyTorch for Deep Learning - DeepLearning.AI
- Deep Learning Specialization - DeepLearning.AI
- Machine Learning Specialization - Stanford/DeepLearning.AI
- Mathematics for Machine Learning - Imperial College London
- Python for Everybody - University of Michigan
- Statistics with Python - University of Michigan