Blog
Notes from the team
Thoughts on building practical AI — language models, retrieval, evaluation, and the products we ship.
-
Why character-based AI is a serious use case for RAG
Character chat is often dismissed as entertainment. We think it's one of the most demanding — and revealing — testbeds for retrieval-augmented generation.
-
Building LogicKor: lessons from evaluating Korean LLMs
Before you can improve a model, you have to measure it honestly. Here's what we learned designing a benchmark that became a standard for Korean LLM evaluation.
-
Introducing Aiocia
A fresh face in the AI scene, fueled by a “why not?” spirit — and a belief that cutting-edge technology should serve people first.