# Your users find the answer in one question

Your users spend more time searching for information than using it. The agent finds the answer in one question.

## What it does

Your users ask a question in natural language. The system searches your documents, validates the results and returns a precise answer with cited sources.

### Natural language

No keywords, no special syntax. The user asks their question as they would to a colleague.

### Cited sources

Each answer references the exact passages from source documents. Verifiable and traceable.

### Zero hallucination

Chain validation to ensure the answer is faithful to the documents. No fabrication.

## The RAG pipeline, step by step

01

### Document ingestion

Your documents are imported, chunked and prepared for indexing.

02

### Semantic indexing

Each passage is transformed into a semantic vector for meaning-based search.

03

### User question

The user asks their question in natural language, in any language.

04

### Semantic search

The most relevant passages are retrieved by semantic similarity.

05

### Chain validation

The LLM verifies the consistency and faithfulness of the answer against the sources.

06

### Sourced answer

The user receives a precise answer with references to the original documents.

## Why it's different

### Chain validation

The system verifies that each claim is supported by a source passage. If it isn't, it says so.

### Sources always cited

Each answer includes exact references: document, page, passage. The user can verify.

### No-code admin

Document management, configuration and monitoring with no technical expertise required.

## All document formats

Import your documents as-is. The pipeline handles extraction and indexing automatically.

- PDF
- Word
- CSV
- OCR Images
- Web pages

## Multilingual search

Ask a question in French, find the answer in an English document. The system understands meaning, not just words.

- Question in one language, documents in another
- Cross-lingual semantic search
- Answer in the language of the question

## Use cases

### Product documentation

Your users find the answer in the product docs without browsing through them manually.

### Internal knowledge base

Procedures, guides, internal FAQ -- accessible with a simple question.

### Regulatory documents

Codes, conventions, legal texts -- find the right article in one question.

### Training materials

Transcribed videos, PDF materials, exercises -- an intelligent training assistant.

## Make your documents accessible

Let's analyze your document search and RAG needs together.

[Discuss your document base](https://www.agora.software/en/contact/)