I build backend systems, web products, and practical AI features with Python and TypeScript.
Technologies I use to build AI-powered solutions and full-stack applications
From AI research to production systems, building solutions that bridge cutting-edge technology with real-world impact
Maintaining PowerCard CMS for European PayTech, collaborating across departments, and driving AI adoption initiatives.
Trained multiple departments in PowerCard usage during system migration
Assisted in writing Business Requirement Documents with BA Department
Facilitating AI adoption across the department through dedicated initiatives and training sessions.
Co-founded Generated By Humans (GBH) with a mission to democratize AI knowledge for the general public. Through engaging content across social platforms, we break down complex AI concepts into accessible, practical insights that empower everyday users to understand and leverage AI technology.
Whether it's you or anyone else, you can get to know about our AI content on YouTube
Designed DB schemas, ran EDA, and built LSTM / Isolation-Forest pipelines for sensor-time-series anomaly detection in industrial IoT.
University of Thessaly
Led crowdfunding-platform capstone project
Designing deep-learning architectures for time-series anomalies; applying results to sensor data and payment-fraud streams.
Maintaining PowerCard CMS for European PayTech, collaborating across departments, and driving AI adoption initiatives.
Trained multiple departments in PowerCard usage during system migration
Assisted in writing Business Requirement Documents with BA Department
Facilitating AI adoption across the department through dedicated initiatives and training sessions.
Co-founded Generated By Humans (GBH) with a mission to democratize AI knowledge for the general public. Through engaging content across social platforms, we break down complex AI concepts into accessible, practical insights that empower everyday users to understand and leverage AI technology.
Whether it's you or anyone else, you can get to know about our AI content on YouTube
Designed DB schemas, ran EDA, and built LSTM / Isolation-Forest pipelines for sensor-time-series anomaly detection in industrial IoT.
University of Thessaly
Led crowdfunding-platform capstone project
Designing deep-learning architectures for time-series anomalies; applying results to sensor data and payment-fraud streams.
A selection of backend systems, product builds, and practical AI projects.
A production-grade FastAPI backend deployed on AWS using ECS Fargate, Terraform, and CI/CD. Includes S3, DynamoDB, IAM, ALB networking, and an AI-ready RAG architecture. Built to understand what actually breaks when moving from localhost to real cloud production.
Official public website for KETAK Research Center. Built with Next.js, focusing on content structure, performance, multilingual support, and maintainability. Designed for a real research organization with non-technical stakeholders.
Frontend platform for educational content delivery. Built with Next.js with emphasis on UX clarity, content hierarchy, and responsive layouts. Designed to scale as a content-driven product.
An event-driven backend system using S3, Lambda, DynamoDB, EventBridge, and DLQs. Focuses on async processing, reliability engineering, and observability. Fully managed with Terraform.
Personal website and blog built with Next.js and Sanity CMS. Used to document real engineering projects and long-form technical writing.
RAG-ready AI SaaS architecture built with Next.js and FastAPI. Designed for production-grade auth, streaming chat, persistence, and a pluggable AI boundary — with inference intentionally stubbed to focus on system correctness and scalability.
