Updates, insights, and behind-the-scenes looks at projects and research.
Just completed the Agentic AI course by DeepLearning.AI and really enjoyed how practical it was. It gave me a much better understanding of agentic design patterns like reflection, tool use, planning, and multi-agent workflows, and how these ideas can move LLMs from simple prompting to more autonomous systems. Excited to keep building and experimenting with these concepts.
I’ve always been interested in geography and languages, and I often wondered: how many people in the world could I actually communicate with based on the languages I speak?
So I built a small side project for fun: Speakrow.
Speakrow.com lets you select the languages you speak and instantly shows your global linguistic reach. It updates an interactive world map to highlight where you can communicate, calculates the share of the world’s population you can reach, gives you a “survival score,” and breaks down your coverage by continent.
Just a fun hobby project combining my interests in languages, data, and maps. Go check it out!
Completed Claude Code in Action, a hands-on course focused on using Anthropic’s Claude for real development workflows.
Learned how to use Claude for code generation, debugging, refactoring, documentation, and structured problem-solving. Covered practical prompt strategies, iterative development with AI, and how to integrate Claude into day-to-day coding tasks to improve speed and code quality.
Strong focus on applying AI as a real coding assistant, not just theory, but practical implementation.
Today I officially start my semester abroad at Korea Advanced Institute of Science and Technology in Seoul, South Korea.
During this exchange, I’ll be taking courses from the MBA programs while also immersing myself in Korean culture, exploring Seoul, and experiencing daily life in such a culturally rich and very different country.
I’m grateful for this opportunity and looking forward to the months ahead.
As part of my MSc in Data Science at Copenhagen Business School, I recently completed a project in Predictive Analytics focused on forecasting daily electricity load in Denmark, a country with one of the highest levels of renewable energy integration. The study applied time series approaches like ARIMA and seasonal ARIMA, as well as dynamic regression models that included solar and wind generation data.
The goal was to explore how these methods perform when renewable energy is part of the system, and to address key questions: How do different forecasting techniques compare? Does including renewable generation improve accuracy? And how does forecast performance change across different time horizons and approaches?
This type of research is becoming increasingly important as energy systems worldwide shift toward renewables. Reliable forecasting helps grid operators balance supply and demand, plan for variability, and ensure stable access to electricity as countries move closer to decarbonization targets.
