Introducing the Aether Blog

We build memory infrastructure for AI agents. Until now, everything we learned along the way lived in internal docs and pull-request threads. This blog is where that work comes out into the open.

What you'll find here

Three kinds of posts, no filler:

• Comparisons — honest, externally verifiable looks at how Aether stacks up against other memory layers, tradeoffs included.
• Engineering — how the engine actually works: vector search, document storage, embeddings, and the decisions behind them.
• Guides — practical, copy-pasteable walkthroughs for giving your agent persistent memory and document RAG.

Memory in five minutes

Most posts come with code you can run. Here's the whole idea in one snippet — scope memory to a user, store a fact, recall it later by meaning:

import { Memory } from "@aether-ai/sdk";
 
const memory = new Memory("user-42", { apiKey: process.env.AETHER_API_KEY });
 
await memory.remember("Prefers concise weekly summaries");
 
const hits = await memory.recall("communication style", { k: 5 });
for (const hit of hits) {
  console.log(hit.score, hit.text); // higher score = more relevant
}

One platform, not three

The reason we started Aether: agent memory usually means stitching together a vector database, an embeddings provider, and object storage — each with its own SDK and its own invoice. Aether bundles them:

One API, one bill, sub-millisecond recall — so your agent fleet can scale without your bill scaling with it.

What's next

We're kicking things off with a deep comparison against other memory layers, then moving into engineering write-ups on how recall stays fast as your data grows.

In the meantime, the fastest way to get a feel for Aether is to try it: spin up an API key, or take the pricing for a quick reality check against what you're paying today.