Corpus

338,701 passages, drawn from 13,387 papers.

A carefully assembled, de-duplicated body of orthopedic biomedical literature, drawn from legitimate open-access sources and split into retrievable passages.

338,701

Passages

13,387

Papers

Specialty · Orthopedics

At a glance

Passages: 338,701 deduplicated passages

Papers: Roughly 13,387 source papers

Specialty: Orthopedic clinical decision support (the current beachhead)

Embeddings: OpenAI text-embedding-3-large

Index: FAISS, built for fast similarity search

Status: Deliberately frozen during validation

How the corpus is built

Ingestion & curation

Papers are pulled from legitimate open-access channels, deduplicated, and split into passages.

Embedding

Each passage is converted into a vector using OpenAI's text-embedding-3-large.

Indexing

Those vectors are stored in a FAISS index built for fast similarity search.

Why the corpus is the moat

AskStein's central bet — confirmed empirically in its own testing — is that retrieval quality, not model size or fine-tuning, is the dominant driver of answer accuracy. Improving the corpus and the retrieval pipeline moves the needle more than swapping in a bigger model. Anyone can call an LLM; the defensible asset is the curated body of evidence and the standards that keep it clean.

See how the corpus is actually used at query time in the documentation →

The corpus is deliberately frozen during this validation window — you don't change what you're validating — and is designed to scale substantially once that window closes.