Remember the useful details.
Keep memory connected to the right person or entity. Carry relevant facts and preferences across conversations without mixing everyone’s context together.
Aether.
A new conversation. A lasting memory.
For developers building agents that need context beyond a single chat. Aether brings durable memory and knowledge retrieval together, so useful facts and source material are ready for the next answer.
Good work shouldn’t start from zero.
Preferences, decisions, and hard-won knowledge belong in the next conversation, too. Aether gives that context a place to live, so your agents can pick up with what matters.
Find the source behind an answer.
Bring documents into your knowledge base and retrieve relevant passages. Give an agent source material to work from when memory alone isn’t enough.
Meet your agents where they work.
Connect through the API, MCP, or SDKs for Python, TypeScript, Go, and .NET. Scoped keys and tenant boundaries keep access tied to the context it needs.
Store a preference. Recall it when it matters.
A small, inspectable example using Aether’s TypeScript SDK. Your application provides the memory, asks for relevant context, and chooses what to use.
Choose the right context.
Use a stable entity ID for the person or context your application has authorized. Keep the API key in your server environment.
Save a useful detail.
Call remember with a supplied preference. It becomes stored context; your application decides what should be remembered.
Retrieve before the next answer.
Call recall with a question and inspect the returned memories. Pass relevant context to your model when it helps.
TypeScript · server-side
import { Memory } from "@aether-ai/sdk";
const memory = new Memory("customer-42", {
apiKey: process.env.AETHER_API_KEY,
});
await memory.remember(
"Prefers concise project updates",
);
const matches = await memory.recall(
"How should I format the next update?",
{ k: 3 },
);
for (const match of matches) {
console.log(match.text);
}Aether returns context, not the final answer. This example prints retrieved text and makes no model call.
Put Aether to work.
Explore all solutionsAI application developers
Build AI applications with persistent memory and source retrieval from Aether. Inspect the user context and documents behind each response.
Consulting firms
Retrieve engagement briefs, discovery notes, and supporting research with Aether. Keep the sources behind consulting recommendations available for review.
Agent memory
Give AI agents persistent, entity-scoped memory with Aether. Remember useful facts across sessions and recall relevant context without replaying every chat.
One part of a bigger picture.
Lasting knowledge. A shared plan. Authorized action. Explore the other products in our family.
Make the next conversation count.
Start with the TypeScript quickstart: store a memory, retrieve relevant context, and decide how your application uses it.