About
What started as a way to explore data science and AI outside the classroom has grown into a collection of projects that showcase what can be built when enjoyment and curiosity meet determination.
With the goal of continuous learning and fun, during our studies back in april 2024 a friend and I co-founded Nordic Raven Solutions, a data consultancy. Honorary collaborations include Paper Check and my University Dormitory.
Since then I have graduated with my Masters in Business Intelligence from Aarhus University (Oct. 2025), and built a portfolio to showcase my determination to apply my skills outside of the classroom too.
What I love doing
| Category | Technologies | What It Means |
|---|---|---|
| Generative AI / LLMs | LangChain, RAG, ReACT, Gemini 1.5, OpenAI GPT-4, QLoRA (Fine-tuning), Transformers, NLP | Specializing in cutting-edge Gen-AI and Agentic Systems to unlock new business value and automation. |
| ML Engineering / MLOps | Docker, Git, FastAPI, LangSmith, Azure, GCP, Railway | Full-Lifecycle MLOps: Adept at building, deploying, monitoring, and scaling models from proof-of-concept to production. |
| Core Data Science / ML | Python, scikit-learn, TensorFlow, XGBoost, FinBERT, Classification/Regression | Core Data Science Competence: Strong foundation in classic and complex models (e.g., XGBoost, FinBERT) for robust predictive analysis. |
| Data Engineering / BI | PostgreSQL, SQL, ETL/ELT, Arelle (XBRL), Apache Superset, Power BI | End-to-End Data Readiness: Proficient in ETL/ELT, infrastructure, and ensuring the data quality required for reliable DS/ML applications. |
Generative AI / LLMs
Technologies: LangChain, RAG, ReACT, Gemini 1.5, OpenAI GPT-4, QLoRA (Fine-tuning), Transformers, NLP
What It Means: Specializing in cutting-edge Gen-AI and Agentic Systems to unlock new business value and automation.
ML Engineering / MLOps
Technologies: Docker, Git, FastAPI, LangSmith, Azure, GCP, Railway
What It Means: Full-Lifecycle MLOps: Adept at building, deploying, monitoring, and scaling models from proof-of-concept to production.
Core Data Science / ML
Technologies: Python, scikit-learn, TensorFlow, XGBoost, FinBERT, Classification/Regression
What It Means: Core Data Science Competence: Strong foundation in classic and complex models (e.g., XGBoost, FinBERT) for robust predictive analysis.
Data Engineering / BI
Technologies: PostgreSQL, SQL, ETL/ELT, Arelle (XBRL), Apache Superset, Power BI
What It Means: End-to-End Data Readiness: Proficient in ETL/ELT, infrastructure, and ensuring the data quality required for reliable DS/ML applications.
My Latest Projects
House Prices Prediction
Real Estate Economics Meets Machine Learning
Achieved top 8.1% performance (rank 476/5,887) in a Kaggle competition by combining hedonic pricing theory with modern ML techniques. Built an 8-model hybrid ensemble with 2-level stacking, demonstrating how domain knowledge guides feature engineering.
CodePractice.AI
Full-Stack AI-Powered Data Tutoring Platform
Built a mobile-first application using Gemini 1.5 to provide intelligent, real-time feedback on user-submitted Python and SQL code. Enables full code execution in the browser without backend dependencies.
AI News Digest
Multi-Agent News Curation System
Developed a multi-agent system (ReACT/RAG) for automated, quality-scored curation of AI research news, complete with LangSmith observability. Features RAG-based deduplication, ReACT agents with tool use, and weekly themed execution.
FinSight
Business Intelligence & ETL Architecture
Created a robust ETL pipeline that extracts, normalizes, and validates financial facts (XBRL) from thousands of SEC/ESEF filings into a PostgreSQL warehouse. Interactive demo enables analysis of any publicly listed company.
Novo Nordisk Analysis
Financial & Competitive Analysis
Strategic analysis of Novo Nordisk with comprehensive financial metrics, peer comparison, and 5-year trend analysis. Dashboard showcases market positioning, financial fundamentals, R&D efficiency, and innovation returns.
CurRag
RAG System for University Notes
Built a Retrieval-Augmented Generation system for querying university lecture notes using LangChain, ChromaDB, and OpenAI. Features semantic search with natural language queries and a Streamlit web interface.
Get in Touch
Interested in working together? Let's talk :)
Contact Information
Feel free to reach out for project collaborations, consulting opportunities, or just to discuss AI/ML and data analytics.