We are an engineering-first company working towards making AI systems work reliably over long horizons by building the necessary data layer.
The way we store knowledge was built for analysts and dashboards, and agents need something different underneath them. Models will keep getting better at reasoning, and the context they can hold will keep getting larger. A larger haystack does not make the needle easier to find. Surfacing the right context is a separate job, and it is not what models are built for. We think that job belongs to a new kind of data layer.
Our open-source python library for creating knowledge repositories that ingest unstructured and structured multi-source data and expose a single query substrate, built for integrating into long-horizon AI agents.
We’re structured as a small, engineering-focused team that operates in the data infrastructure space.
We’re dedicated to productizing and researching how to provide the right data at the right moment for long-running agents, which seems like a trivial problem but is actually much harder once you go down the rabbit hole.
Hence why we’re called Deyta (/ˈdeɪ.tə/).
A piece of AI slop we came across, and it sums up why we’re doing this:
“AGI won’t be achieved on top of postgres & vector DBs.”
