What can I build?

Real things you can build with Aether, explained without jargon.

Aether stores your documents and lets AI find the right ones when someone asks a question. Here are concrete things people build with it.

AI chatbot for your docs

The problem: You have help docs, a wiki, or internal guides and people keep asking the same questions. A generic chatbot hallucinates or gives outdated answers.

What Aether does: You feed your docs into Aether. When someone asks a question, Aether finds the 3-5 most relevant sections and hands them to an LLM like Claude. The LLM answers based on your actual docs, not its training data.

Who builds this: Developer advocates, support teams, SaaS companies.

Start here: Zero-to-chatbot guide — build one in 30 lines of Python.

Research assistant that cites sources

The problem: You're working with a large collection of papers, reports, or legal documents. You need answers with citations pointing back to the original source.

What Aether does: Each chunk Aether returns includes a doc_id and ranking score. You tell the LLM to cite the source document, so every answer links back to the original. No more "where did the AI get that?"

Who builds this: Researchers, analysts, law firms, students.

Start here: RAG quickstart — shows the retrieve-and-cite pattern.

Customer support bot

The problem: Your support team answers the same questions over and over. You want an AI that handles the easy ones automatically, using your actual knowledge base.

What Aether does: Load your support articles and FAQ into Aether. When a customer asks a question, retrieve the relevant articles and generate an answer. If confidence is low (high search score), escalate to a human.

Who builds this: Support teams, e-commerce companies, B2B SaaS.

Start here: Integrations overview — pick your LLM provider and wire it up.

Personal knowledge base / second brain

The problem: You save notes, bookmarks, articles, and PDFs everywhere. When you need something, you can't find it.

What Aether does: Dump everything into Aether — notes, PDFs, articles, meeting transcripts. When you search "that article about pricing strategy from last month," Aether finds it by meaning, not keywords.

Who builds this: Writers, founders, anyone with too many notes.

Start here: CLI reference — use aether-py insert to load files from the command line.

AI tutor grounded in course material

The problem: Students ask an AI for help but it gives generic answers that don't match the course material, textbook, or professor's approach.

What Aether does: Upload the syllabus, lecture notes, and textbook excerpts. When a student asks a question, Aether pulls the relevant course material and the AI answers in the context of this specific class.

Who builds this: Educators, edtech platforms, tutoring services.

Start here: Zero-to-chatbot guide — swap in your course materials.

Not sure? Start with the 30-line chatbot

If you're not sure where to begin, the zero-to-chatbot guide gets you a working prototype in minutes. Once you see it working, you'll know what to build next.