Hi, I am
AI Engineer
AI & Machine Learning Engineer with 3+ years of experience designing and deploying intelligent, data-driven systems across agronomy, energy management, biomedicine, and public sectors. At PwC, I develop and scale AI solutions that combine Generative AI, multi-agent systems, and MLOps best practices to deliver measurable business outcomes. My experience spans research and production engineering — from leading the ImageCLEFmedical 2022 Caption Prediction Task (1st & 2nd place) to building robust, production-grade ML pipelines and AI applications. Skilled in deep learning, NLP, and Generative AI, I leverage Python, SQL, Docker, and Azure to create scalable systems that translate innovation into real-world impact.
Experience
AI Engineer
I design and deploy production-ready multi-agent AI systems and RAG-based solutions across multiple business use cases, enabling context-aware interactions and scalable decision-making.
I architect and implement cloud-native ML orchestration platforms with asynchronous execution, autoscaling, and experiment tracking, ensuring reliable end-to-end AI workflows.
I build end-to-end machine learning pipelines, from data processing and AutoML model training to deployment and lifecycle management, delivering secure and scalable solutions for enterprise and public-sector applications.
I develop and optimize ML-powered REST APIs, enhancing performance and accelerating analytics and operational pipelines while following MLOps and DevOps best practices.
I lead cloud solution delivery, mentor engineers, and collaborate with cross-functional teams to translate complex AI concepts into actionable solutions aligned with business goals.
I also lead the PwC Greece Python Community, promoting clean architecture, best practices, and a culture of continuous learning.
Apr 2024 - Current
Software Engineer
@ PPC S.A.
Served as the sole Java Spring developer for the Energy Trading Unit at PPC, leading the design and implementation of ETL pipelines.
Developed decision-making dashboards for 100+ users using Primefaces, Plotly, and CSS.
Feb 2023 - Apr 2024
IT Support Engineer
@ Ministry of National Defense, Hellinc Army, Greece
Served as an IT Support Engineer, providing systems support and IT consulting.
Dec 2022 - Nov 2023
Data Science Intern
@ Augmenta
Led agricultural optimization project with scikit-learn and OpenCV, creating tools to enhance logistics and mitigate crop damage for 250+ farmers.
Apr 2022 - Oct 2022
Junior Machine Learning Researcher
Conducted research on NLP, Computer Vision, and Deep Learning for medical diagnosis, using PyTorch, TensorFlow, and scikit-learn for model development and evaluation.
Led the annual ImageCLEFmedical Caption Task campaign, achieving 1st and 2nd place in the main evaluations among 10 teams.
Oct 2021 - Oct 2022
Education
Research and publication
AUEB NLP Group at ImageCLEFmedical Caption 2022
Placed 1st and 2nd in ImageCLEFmedical, an international
competitive campaign for image analysis and retrieval.
Researched Medical Image Captioning using CNNs, RNNs, Transformers with PyTorch and
TensorFlow.
2022
Certifications
Projects
Exploring Uni-modal, Cross-modal, and Multi-modal Diagnostic Captioning
B.Sc. Thesis Deep Learning & NLP research on Medical Image Captioning
- Python
- TensorFlow
- PyTorch
- Numpy
- Pandas
- NLTK
- Huggingface
- matplotlib
PumpRAG
PumpRAG is an innovative Generative AI agent utilizing a Retrieval-Augmented Generation (RAG) architecture to enhance the gym workout experience. This advanced system combines the power of natural language processing with a robust retrieval mechanism to provide precise and contextually relevant responses to fitness-related queries.
- Python
- PyTorch
- OpenAI
- Azure
- Pinecone
- FastAPI
Overfit Level I
A Multi-Class Prediction of Obesity Risk Machine Learning Application.
- Python
- scikit-learn
- xgboost
- Pandas
- seaborn
- SHAP
- Flask
- Docker
- Huggingface Spaces
Tournament Management Android App
A management app for amateur football & basketball tournaments, developed in Java. Used modern practices such as the Model‑View‑Presenter pattern, MemoryDAO and JUnit Tests.
- Java
- Draw.io
- JUnit
- MVC
- Android
- UML
- XML
Tika Toka App
A video streaming app for Android. It follows the Publisher-Consumer model, where the creator uploads videos, which can be viewed by the subscribers (similar to TikTok).
- Java
- Kotlin
- JUnit
- MVC
- Android
- Threads
- Sockets
Espresso machine app
In this project we tried to make a new updated interface of a typical non-smart home espresso machine. We added new functions, which current one does not have. Moreover, this application addresses the visually impaired people, with voice assistant.
- Java
- Kotlin
- JUnit
- Android
- UML
- XML
Skills
Programming
- Python
- Java
- HTML/CSS
- R
- Kotlin
Machine Learning & AI
- MLFlow
- FLAML
- AutoML
- LangGraph/LangChain
- TensorFlow
- PyTorch
- scikit-learn
- OpenCV
- Conda
- NLTK
- scipy
- HuggingFace
- Numpy
- Pandas
- matplotlib
- seaborn
Database
- MS SQL
- Oracle SQL Developer
- SQL Workbench
- MongoDB
- PostgreSQL
Tools
- Azure
- Fabric
- Docker
- DevContainers
- Linux
- Git
- Poetry/uv
- Azure DevOps Pipelines
- GitHub
- Latex
- Android Studio
Frameworks
- FastAPI
- Django
- Spring Boot
- Primefaces
- JavaServer Faces (JSF)
- Flask
- Selenium
Contact
Hit me up on LinkedIn or send me a direct email