Blog
The Kortexio Blog
Notes on LLM memory, agentic tooling, and building AI infrastructure in Europe.
31 July 2026 · 5 min read
Kortexio is now OpenAI-compatible — for real
POST /v1/chat/completions, /v1/models, and /v1/embeddings — drop-in OpenAI SDK support with ambient memory, while /api/chat stays Ollama-compatible.
Read article →10 July 2026 · 8 min read
Keeping AI infrastructure inside the EU — and why it's getting harder
A practical look at data residency, GDPR, and what EU-hosted by default really requires under the hood.
Read article →5 July 2026 · 7 min read
Structured memory vs. raw embeddings: what your model actually needs
Why a compiled, evolving session memory outperforms a pile of similarity-searched chunks for most conversational products.
Read article →1 July 2026 · 6 min read
Why we made LLM memory a URL, not an SDK
The engineering argument against rebuilding a vector store for every app — and how a proxy-shaped memory layer changes integration cost from weeks to minutes.
Read article →