Skip to main content

​
What EDN is

EDN is provenance-aware external memory middleware for LLM systems, built by TensorPRO. It provides structured persistence, typed correction channels, and deterministic retrieval for the context your application feeds to a language model. You bring your own model. EDN never supplies a model and never routes requests to one. It stores what your systems and users tell it, classifies every record by where it came from, and returns grounded, cited context on demand. Which model consumes that context is entirely your choice.

​
The Platform and its two products

EDN is delivered through a single platform, the Console, at console.tensorpro.ai. Two products sit inside it:

Workspace

A chat workspace grounded in EDN memory. You connect your own model, upload sources, and get answers with provenance and citations.

API

The EDN API for building memory into your own applications: write records, correct them, and retrieve governed context.
Access to the Console and its products is tier based. A new account starts on a free trial, with Pro, Team, and Enterprise tiers available. See Plans and tiers.

​
Base URL

This base URL is the EDN evaluation endpoint. It is intended for evaluation and testing. Do not describe it or rely on it as a production endpoint.

​
Where to go next

Quickstart

Mint an API key and make your first authenticated call.

Authentication

How EDN authenticates requests with an API key.

User Guide

Using the Console, Workspace, and your account.

Developer Docs

How EDN works and how to build against it.