🚀 Projects Portfolio

A curated collection of projects spanning data engineering, analytics platforms, cloud architectures, and full-stack applications.

I focus on production-oriented design: scalability, reliability, observability, and clear data modeling — from real-time streaming systems to serverless pipelines and interactive data products.

📈 Data Platform for Detecting Trading Trends in Financial Tick Data

Apache Flink · Kafka · InfluxDB · Grafana · Python · Java · Docker

Impact:
Real-time analytics on high-volume financial tick data, generating trend indicators and trading signals in seconds instead of hours.

What I built

• Stream ingestion using Apache Kafka with event-time processing.
• Low-latency analytics and windowed computations in Apache Flink.
• Time-series persistence in InfluxDB for fast queries and retention control.
• Interactive Grafana dashboards for monitoring trends and indicators.

🔗 Links:
GitHub ·
Technical Report

📦 Parcel Delivery Quality Data Pipeline

Azure Databricks · Delta Lake · Delta CDF · Apache Spark · Python · Medallion Architecture

Impact:
Scalable, incremental analytics pipeline enabling reliable KPI reporting for parcel delivery quality monitoring.

What I built

• End-to-end data pipeline using Bronze–Silver–Gold architecture.
• Incremental data processing with Delta Change Data Feed (CDF).
• Optimized Gold-layer tables for performance and analytical workloads.
• Checkpointing and fault-tolerant Spark jobs on Azure Databricks.

🔗 Links:
GitHub

☁️ Serverless AWS Data Pipeline Architecture

AWS CDK · Lambda · Glue · Step Functions · S3 · Athena · Python · SQL

Impact:
Fully serverless analytics pipeline with automatic scaling, fault tolerance, and pay-per-use cost efficiency.

What I built

• Infrastructure as Code using AWS CDK.
• Event-driven ingestion and validation with AWS Lambda.
• Data transformation via AWS Glue, orchestrated by Step Functions.
• Analytics and reporting using Amazon Athena over S3 data lakes.

🖼️ Architecture 🔗 Links:
GitHub

🗂️ Kanban Board Application

Scala · ScalaFX · UML · Unit Testing · JSON · Git

Impact:
Desktop task-management application inspired by Kanban principles for visual workflow tracking.

What I built

• Kanban board UI using ScalaFX.
• Task state management with persistent JSON storage.
• UML-driven design and unit-tested core logic.
• Emphasis on usability and real-time task updates.

🖼️ UI Preview

🔗 Links:
GitHub

🌍 Climate Change Communication Website

Python · Data Analysis · Visualization · Dash · Heroku

Impact:
Interactive data product communicating greenhouse gas emission trends to Nordic audiences.

What I built

• Data exploration and cleaning of global emissions datasets.
• Interactive visualizations showing GDP per capita vs CO₂ emissions.
• Dash-based web application deployed on Heroku.
• Collaboration in a 5-person team as part of a university data science project.

🖼️ Visualization

🔗 Links:
Live Website

🧠 Quiz Application (Full-Stack Web App)

JavaScript · HTML · CSS · Deno · Oak · PostgreSQL · Heroku

Impact:
Three-tier web application for creating and answering multiple-choice quizzes via UI and API.

What I built

• REST API using Deno + Oak.
• PostgreSQL-backed persistence layer.
• Client-side UI for quiz creation and participation.
• Deployed full stack on Heroku.

🖼️ Application UI

🔗 Links:
GitHub ·
Live Website

💉 Vaccination Database Design

SQL · Python · UML · Data Normalization · Relational Databases

Impact:
Normalized relational database system for managing vaccination, patient, and doctor records.

What I built

• Conceptual and logical design using UML.
• Data normalization and schema optimization.
• Data ingestion and transformation with Python.
• Team collaboration (4 members) for database design and implementation.

🖼️ Database Design

🔗 Links:
GitHub

🧬 Machine Learning for Personalized Treatment Decision-Making

Causal Inference · Reinforcement Learning · Healthcare Analytics · Research

Impact:
Bachelor’s thesis analyzing machine-learning approaches for personalized and dynamic treatment strategies in healthcare.

Focus

• Individualized Treatment Effect (ITE) estimation via causal inference.
• Reinforcement learning for optimal dynamic treatment regimes.
• Comparison of objectives, data sources, evaluation methods, and limitations.
• Discussion on interpretability, trust, and hybrid causal-RL approaches.

📄 Thesis:
PDF on Google Drive

📫 Contact

If you’d like to discuss any of these projects, collaboration opportunities, or research ideas, feel free to reach out.