Making the
world’s data
AI-native.
Intelligence needs context.
We build the infrastructure that connects the two.
Our flagship product
Built by EvokoaPostgreSQL. Connected to intelligence.
Your data.
A world of context.
Polygres is our solution for AI retrieval. It brings graph, vector, and full-text search together as a virtual layer over PostgreSQL, helping applications find the context they need.
Explore the platform, meet the open-source engines, and start building.
Explore polygres.comThe company
Better context.
Greater possibilities.
Evokoa is the research and engineering company behind Polygres.
We believe the next step for AI depends on how well it understands the world around it. That starts with data: the facts, relationships, and knowledge that give an answer its meaning.
We’re building the infrastructure to bring that context within reach. Our work brings together databases, search, and open-source engineering.
Start with the data.
The knowledge is already there.
Make it useful to AI.
Keep the connections.
Meaning lives in relationships,
as well as individual records.
Build in the open.
Put the foundations in the hands
of the people building on them.
Open by design
Good foundations
belong to everyone.
We develop the PostgreSQL extensions behind Polygres in the open. Read the source, explore the ideas, and help shape what comes next.
Explore our GitHubA little more context
About Evokoa.
What is Evokoa?
Evokoa is a research and engineering company based in San Francisco. We build infrastructure that helps AI applications find and use relevant data, with a focus on PostgreSQL, retrieval, and context.
How are Evokoa and Polygres related?
Evokoa is the company behind Polygres. Polygres is our solution for AI retrieval, bringing graph, vector, and full-text search together as a virtual layer over PostgreSQL. Visit polygres.com for the product, documentation, and ways to start building.
What does Evokoa build in the open?
We develop pgGraph and pgContext, open-source PostgreSQL extensions for graph queries and AI search. Their source code and technical documentation are available through the Evokoa GitHub organization.