Eemeli Kuittinen

Our team built a service that transcribes audio files and YouTube videos, then summarizes the transcripts with a language model.

I was the Scrum Master and lead developer. I led the sprints, chose the tools, and handled much of the coding and testing. We deployed the interface in containers on CSC Rahti 2 and used Puhti HPC for processing.

KAMK · Digital Services project course

Our team used computer vision to identify culvert pipe materials from photographs.

I augmented the training data and checked annotations in Label Studio. I also helped train and fine-tune the YOLO model.

KAMK · Applying Artificial Intelligence project course

I built a pipeline that collects Finland's railway traffic data through the FinTraffic API.

The pipeline uses dbt and DuckDB to model the data, then displays it in Evidence. I documented how the data moves through the pipeline and wrote automated checks to validate the results.

Data Platform project