Macha Concepts

Data Sources

Définition

Data sources (also called knowledge sources) are the knowledge an agent reads from — uploaded documents, indexed websites, and live content from connected apps like Google Docs, Notion, and Confluence. Where connectors give an agent actions, data sources give it context.

Aussi connu sous : knowledge sourcesknowledge baseagent knowledge

How it works

You add sources on the Sources page: upload files (PDF, DOCX, CSV, XLSX, TXT), crawl an entire website, add a single web page, or connect live services. Small spreadsheets (≤2,000 rows) are injected straight into the agent's context; larger files and websites are chunked and embedded so the agent searches them via its search_knowledge and get_document tools; live connector sources (Google Docs, Notion) are read fresh every time so the agent always sees the current version.

When you attach a source to an agent you can scope it to all documents or only selected ones. Adding a Google or Notion source auto-links the read tools the agent needs.

Static vs live — which to pick

The choice comes down to one question: does the content change? Static content (policies, FAQs, manuals, Help Center articles) is best uploaded or indexed once. Dynamic content (pricing sheets, on-call rosters, anything that changes weekly) is best added as a live connector source so the agent reads the latest version instead of a stale snapshot.

This is Macha's application of retrieval-augmented generation: the agent grounds its answers in your documents rather than guessing from training data.

Questions fréquentes

Are "data sources" and "knowledge sources" the same thing?

Yes — they're two names for the same feature. The Sources page labels them knowledge sources; the concept is identical.

How is a data source different from a connector?

A connector lets an agent take actions in an external tool; a data source gives the agent knowledge to read. Many agents use both at once.

Data Sources docs

Termes liés

Connectors

Connectors are Macha's built-in integrations.

Retrieval-Augmented Generation (RAG)

Retrieval-augmented generation (RAG) is a technique where an AI model retrieves relevant information from your own knowledge sources at query time and uses it to ground its answer, instead of relying only on what it memorized during training..

Knowledge Base

A knowledge base is a structured, searchable library of articles — how-tos, FAQs, troubleshooting guides, and policies — that lets customers or agents find answers without contacting support directly..

Grounding

Grounding is the practice of tying an AI model's answers to verified source material — your documentation, live data, or knowledge base — so responses reflect real facts rather than the model's own guesses..

Passez à l'action

Macha est une couche d'agents IA qui se pose sur le help desk que vous utilisez déjà, Zendesk, Freshdesk, Front, Intercom ou Gorgias.

Démarrer l'essai