About
I’m Ricardo Ledan — an AI systems engineer and researcher working across artificial intelligence, distributed systems, and digital archives.
I’m the founder of studio1804, an independent AI research lab, and the creator of rasin.ai, an AI-assisted archival research platform for Haitian history.
My work focuses on building systems where standard approaches break — knowledge graphs for archives, multi-agent systems for document analysis, and self-hosted AI infrastructure that doesn’t depend entirely on the cloud.
I’m particularly interested in the systems layer of AI: how knowledge is structured, how models interact with data over time, and what it takes to make these systems reliable, interpretable, and usable outside of controlled environments.
A lot of that interest comes back to a few core questions:
How is knowledge preserved?
Who owns the infrastructure it depends on?
And how do technical systems shape what people are able to access, understand, and trust?
I write here about the systems I’m building, the technical problems I’m working through, and the ideas that inform them.
Lately, I’ve been reading What Is Intelligence? by Blaise Agüera y Arcas — which reframes intelligence as fundamentally about prediction across biological and artificial systems — and Machine Decision Is Not Final, edited by Benjamin H. Bratton, which situates AI within broader political and cybernetic traditions beyond the Silicon Valley lens. I’ve also been working through Build a Large Language Model (From Scratch) by Sebastian Raschka, which has been useful for grounding my understanding of model architecture from tokenization through fine-tuning.