Juan Buhagiar Duijst

Software Architect & Technical Leader

Designing scalable, event-driven systems for financial trading platforms. Leading high-performance engineering teams. Driving innovation through architecture and DevOps excellence.

View My Work LinkedIn Profile

About Me

Software Architect with Financial Excellence

I'm a software architect and technical leader with 10+ years of experience building scalable, event-driven systems in fintech and trading platforms. I specialize in leading high-performance teams, designing microservices architectures, and optimizing cloud infrastructure.

At ETPA, I scaled a trading team from 5 to 14+ members while migrating from monolithic to microservices architecture, achieving 70% latency reduction and 30% cloud cost savings.

I'm passionate about clean architecture, DevOps excellence, and sustainable community impact. Recognized as one of 500 most ambitious young talents in technology by The Next Web (2017, 2019).

Current Role: Based in Malta, preferring Malta or Remote positions. Vice President at Rota, driving sustainable mobility initiatives.

14+
Team Members Led
70%
Latency Reduction
30%
Cloud Cost Savings
10+
Years Experience

Technical Skills & Expertise

🏗️ Architecture

  • Event-driven systems
  • Microservices design
  • CQRS patterns
  • Event sourcing
  • Monolith migration
  • Scalable systems
  • High availability

☁️ Cloud & DevOps

  • AWS (EC2, ECS, EKS, RDS)
  • Kubernetes clustering
  • Docker containerization
  • CloudFormation IaC
  • Ansible automation
  • GitLab CI/CD pipelines
  • Datadog monitoring

💻 Programming & Frameworks

  • Java & Kotlin
  • Spring Framework
  • Spring Boot
  • Axon Framework
  • MongoDB
  • PostgreSQL
  • Apache Kafka

🎯 Leadership & Process

  • Team scaling & mentoring
  • Technical roadmaps
  • Scrum & Kanban
  • Cross-team collaboration
  • Architecture boards
  • Process optimization
  • Technical decision making

📊 Performance & Optimization

  • Performance tuning
  • System reliability
  • Distributed tracing
  • Cost optimization
  • Disaster recovery
  • Load balancing
  • Caching strategies

🌍 Community & Beyond

  • Open source contribution
  • Workshop organization
  • Sustainable mobility
  • International outreach
  • Technical writing
  • AI/ML fundamentals
  • Computer vision

Professional Experience

Software Development Lead & Architect
ETPA – Trading Team
Nov 2021 – Present (4+ years)
Software Developer
ETPA – Platform
Jul 2017 – Nov 2021 (4+ years)
Computer Vision Engineer
Bright River
Oct 2019 – Dec 2019
Additional Experience
Various Companies
2014 – 2019

Key Projects & Achievements

Energy Trading Platform Modernization
Lead Architect & Technical Lead
Led the migration of ETPA's energy trading platform from event-driven monolith to scalable microservices architecture, enabling seamless team and functionality expansion.
Impact: 70% latency reduction (35ms→10ms), 30% cost savings, 10x scalability, team growth from 5 to 14+ members
Kubernetes AWS Event Sourcing Microservices GitLab CI/CD Datadog
Multi-Repo CI/CD Pipeline Infrastructure
DevOps Architect
Designed and implemented a comprehensive multi-repository GitLab CI/CD pipeline system supporting automated testing, deployment, and distributed tracing across the organization.
Impact: Automated deployments, end-to-end tracing with Datadog, reduced deployment time, improved team velocity
GitLab CI/CD Docker Kubernetes Automation Infrastructure as Code
Regional Expansion to Malta
Team Lead & Recruitment Lead
Successfully established ETPA's Malta branch, recruiting and onboarding 16+ team members (DevOps, developers, QA) while maintaining technical excellence and team culture.
Impact: Successful team expansion, 16+ new hires, maintained quality standards, enabled platform scaling
Leadership Recruitment Team Building Mentoring
Cloud Infrastructure Optimization
Cloud Architect
Optimized AWS cloud infrastructure through resource analysis, on-demand Kubernetes clusters, and usage pattern analysis, resulting in significant cost reduction.
Impact: 30% annual cost reduction, improved resource utilization, maintained performance SLAs
AWS Cost Optimization Kubernetes CloudFormation Ansible
Architecture Board Leadership
Principal Architect
Chaired and led the architecture board overseeing technical decisions and system design across multiple teams, ensuring alignment with business goals and technical excellence.
Impact: Consistent architecture patterns, reduced technical debt, improved cross-team collaboration
Architecture Leadership Decision Making Technical Strategy
International Community Engagement
Volunteer Developer & Workshop Organizer
Organized and conducted technical workshops at international conferences across 8+ countries (South Korea, Ireland, Poland, Germany, Croatia, Sweden, Netherlands, North Macedonia).
Impact: International reach, knowledge sharing, community building, recognized by The Next Web T500
Community Teaching International Knowledge Sharing
XBID Migration & Intraday Trading Platform
Technical Lead
Organised and lead a team of devops and developers to create a new product for the copmany from design, development, testing and deploying MVP. Later on launched the product to our clients. Continuted developing the product to become a big contributer to the revenue stream of the company.
Impact: Product launch, Domain Driven Design, New Revenue stream
Domain Driven Design Architecture Leadership
AI/Computer Vision Publication
Researcher
Published Bachelor's thesis on automatic classification of images from wearable cameras using transfer learning and ensemble methods. Developed methods using pre-trained CNNs (VGG16) with transfer learning for visual lifelogging applications.
Impact: Published research
Computer Vision Transfer learning Publication
Deep Learning
Researcher
Master's thesis on action recognition and temporal localization using I3D networks. Developed 3 novel temporal localization architectures: ActionVLAD, Non-Local Blocks, and VideoGraphs for per-frame action localization in untrimmed videos.
  • 14.31% mAP on Charades dataset (spatio-temporal features)
  • 10.73% mAP on temporal-only features
  • 3 novel architectures for temporal localization
  • Deep analysis across 157 action classes
  • Semantic representation visualization
Computer Vision Deep learning Action Recognition
Data Augmentation for Face Recognition
Researcher
Explored generative approaches and augmentation techniques to improve neural network accuracy with limited training data. Investigated methods beyond simple flipping/rotation to generate new images and improve face recognition model reliability.
Technologies: Deep Learning, Face Recognition, Data Augmentation, Neural Networks
Machine learning GANs Face Recognition Deep learning
Visual Question Answering System
Researcher
Developed two models (CBOW and LSTM) for Visual Question Answering. Given an image and a question, the models predict accurate answers by concatenating precomputed image features with word token outputs. LSTM model showed superior performance by capturing long-term dependencies.
NLP LSTM Deep learning
Teaching a Chatbot with External Knowledge
Researcher
Dialogue model incorporating external knowledge and movie embeddings to overcome common sequence-to-sequence problems (factual errors, repetition, generic responses). Results showed that including similar dialogs helps generate more human-like responses.
NLP Dialogue Modeling Knowledge Integration Deep learning

Education & Recognition

Academic Credentials

Master of Science (MS) in Artificial Intelligence
University of Amsterdam
2017 – 2019
Thesis: "Temporal Localization of Actions in Untrimmed Videos"
  • Published research on action recognition and temporal localization
  • Developed 3 novel architectures: ActionVLAD, Non-Local Blocks, VideoGraphs
  • Achieved 14.31% mAP on Charades dataset with I3D networks
  • Deep analysis of model performance across 157 action classes
Bachelor of Science (BSc) in Artificial Intelligence (2:1)
University of Malta
2014 – 2017
Thesis: "Automatic Segmentation of Indoor and Outdoor Scenes from Visual Lifelogging"
  • Published research on transfer learning for scene classification
  • Achieved 98.28% accuracy using VGG16 and ensemble methods
  • Developed Android application for real-time activity monitoring
  • Erasmus+ Scholar at University of Groningen
Exchange Semester in Artificial Intelligence
University of Groningen
2017

🏆 Awards & Recognition

The Next Web T500

Featured as one of 500 most ambitious young talents in technology

2017 & 2019

🌱 Community Leadership

Vice President, Rota
Sustainable Mobility Advocacy & Community Projects
March 2023 – Present

Let's Connect

I'm always interested in discussing innovative projects, architecture challenges, or opportunities to contribute to high-impact teams. Feel free to reach out!

LinkedIn Email Me