Johann Fernando Medina Paguay
Desarrollador Senior Java & Especialista en Machine Learning / Senior Java & ML Engineer
Senior Java Backend & Machine Learning Engineer (Fintech & Cloud Solutions)
Guayaquil, Ecuador | Phone: - | Email:-
Professional Summary
Senior Software Engineer with 12+ years of experience building scalable backend systems
and deploying applied machine learning solutions for banking and fintech platforms.
Expert in Java (Spring Boot, Quarkus), Python (FastAPI, scikit-learn, TensorFlow), and
cloud technologies (Docker, AWS, Azure). Proven record of improving API performance
(30% latency reduction), automating CI/CD pipelines (25% faster releases), and delivering
production-ready ML models for risk scoring, image classification, and customer
analytics.
Core Skills
Programming Languages: Java (JDK 8–17), Python, Go, SQL, PL/SQL, C++
Frameworks & Platforms: Spring Boot, Quarkus, FastAPI, Hibernate, TensorFlow, PyTorch,
scikit-learn
Cloud & DevOps: Docker, AWS, Azure, Git, Bash, PowerShell, MLflow, CI/CD pipelines
Databases: Sybase, Oracle, PostgreSQL
Other Tools: COBIS CTS, IBM WebSphere, JBoss, REST/SOAP APIs, Microservices,
Agile/Scrum, Linux (RHEL/SUSE)
Professional Experience
Machine Learning Engineer – Anyone AI (Remote)
Jun 2024 – Present
•
Deployed a CNN model (82% accuracy) for vehicle image classification via FastAPI +
Docker on AWS.
•
Built a credit risk scoring pipeline (ROC AUC 0.72) using XGBoost, Pandas, and
MLflow for model versioning.
•
Automated sentiment analysis API using Python + FastAPI to classify customer
feedback in real time.
•
Collaborated remotely with international teams applying Agile methodologies
(Scrum) and CI/CD best practices.
Senior Software Developer – DSJT, Panama
Apr 2013 – Present
•
Designed and integrated enterprise APIs for COBIS core banking (CTS 3.4 & 4.6)
using Spring Boot and Quarkus.
•
Led the migration of legacy components to Quarkus, improving system response
time by 25%.
•
Created CI/CD pipelines with Git and Azure DevOps, reducing deployment time by
25% and improving reliability.
•
Mentored junior developers and standardized documentation, reducing onboarding
time by 40%.
•
Integrated REST/SOAP services and MQ messaging to support high-availability
financial systems.
Software Developer – Business Solutions Tecnología LTDA
Apr 2011 – Dec 2012
•
Developed and maintained enterprise modules for financial institutions using Java
EE and Oracle.
•
Migrated legacy HCRUD modules to Java components (JDK 1.7, Spring 3.7),
improving modularity and scalability.
•
Optimized batch execution by 30%, enhancing system performance and reliability.
Selected Projects
•
Credit-Risk Scoring Model (FastAPI + MLflow + Docker): Processed 350K+
financial records; achieved ROC AUC > 0.72.
•
Vehicle Recognition API (AWS + Docker): Built and deployed CNN image classifier
(82% accuracy).
•
Sentiment Analysis API (FastAPI + Python): Automated classification of customer
reviews; improved response time by 35%.
Education
Escuela Superior Politécnica del Litoral (ESPOL) – Guayaquil, Ecuador
B.S. in Software Analysis and Development -)
Certifications & Training
•
Machine Learning Bootcamp – Anyone AI (2025)
•
Docker – LinkedIn Learning (2024)
•
Kubernetes – Mitocode (2023)
•
Spring Boot – LinkedIn Learning (2023)
•
Full Stack Java & Angular Material – Mitocode (2023)
Languages
•
Spanish: Native
•
English: B2 (Intermediate, improving toward C1)
Career Interests
Backend engineering, fintech integrations, machine learning deployment, and scalable
cloud-native applications.
