Stellar

Knowledge Base

Your docs become your agent's voice

Upload anything your team already has. Stellar reads it, indexes it, and makes it available to your agent in real time. No training step. No reformatting.

product-manual.pdf
2.4 MB · 32 pages
Chunked
1,536-dim embeddings
Indexed in pgvector · ready to query

Upload anything

Five formats, all processed the same way. Drag and drop or paste a URL.

.pdf

Product guides, whitepapers, manuals

.docx

SOPs, scripts, training documents

.txt

Plain text notes and transcripts

.csv

FAQ lists, pricing tables, contact data

.url

Web pages, help docs, knowledge bases

What happens next

01

Chunked

Your document is split into overlapping segments. Each chunk is small enough for precise retrieval but large enough to carry context.

02

Embedded

Every chunk is converted into a high-dimensional vector that captures meaning, not just keyword matches.

03

Searchable

Vectors are stored in pgvector with HNSW indexing. When your agent needs an answer, it retrieves the most relevant chunks in milliseconds.

product-manual.pdf
2.4 MB · 32 pages
Chunked
1,536-dim embeddings
Indexed in pgvector · ready to query

Test it before you go live

The built-in Q&A panel lets you ask questions against your knowledge base and see exactly which chunks the agent retrieves. Fix gaps before your first call, not after.

Your agent should sound like your team

Upload your docs and test a call. It takes about two minutes.