AI Knowledge Answer (RAG)
Answer a question grounded in your own documents. The command chunks the sources, retrieves the most relevant passages (offline TF-IDF) and asks the LLM to answer using only those passages — reducing hallucination.
Overview
AI Knowledge Answer (RAG) is a no-code automation step in the AI category of AutoFlowRPA. It performs its action in a single step when the workflow reaches it. You can configure 10 parameter(s) to control exactly how it behaves. It is one of 369 commands you can drag into a workflow in the visual builder — no scripting required.
What it does
Answer a question grounded in your own documents. The command chunks the sources, retrieves the most relevant passages (offline TF-IDF) and asks the LLM to answer using only those passages — reducing hallucination.
Parameters
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
| API base URL | string | Required | http://localhost:11434/v1 | OpenAI-compatible base. Default = local Ollama (offline). Supports variables |
| API key (blank for local) | string | Optional | Bearer token for cloud providers. Supports variables | |
| Model | string | Required | llama3 | Example: llama3 Supports variables |
| Knowledge sources | multiline | Required | Paste text, or give a file, a folder, or a glob (*.md/*.txt/*.csv/*.json/*.html). Multiple files are all indexed. Example: C:/docs/handbook Supports variables | |
| Question | multiline | Required | Example: What is our refund policy? Supports variables | |
| Passages to retrieve | number | Optional | 4 | How many top passages to feed the model as context. |
| Temperature | number | Optional | 0.1 | |
| Timeout (seconds) | number | Optional | 180 | |
| Store answer in | variable | Required | Example: answer | |
| Store retrieved context in (optional) | variable | Optional | Example: sources |
How to use AI Knowledge Answer (RAG)
- 1
Open AutoFlowRPA and drag the AI Knowledge Answer (RAG) command from the AI group onto the canvas.
- 2
Set its parameters in the side panel — AI Knowledge Answer (RAG) has 10 configurable field(s).
- 3
Connect it into your workflow and press Run. Everything executes locally on your machine.
Frequently asked questions
What does the AI Knowledge Answer (RAG) command do?
Answer a question grounded in your own documents. The command chunks the sources, retrieves the most relevant passages (offline TF-IDF) and asks the LLM to answer using only those passages — reducing hallucination.
Which group does the AI Knowledge Answer (RAG) command belong to?
AI Knowledge Answer (RAG) is part of the AI group. AutoFlowRPA ships with 369+ ready-made commands you can combine without code.
Does the AI Knowledge Answer (RAG) command take any parameters?
Yes — AI Knowledge Answer (RAG) has 10 configurable field(s), all editable in the visual node editor.
More AI commands
Ready to automate your busywork?
Start your free 30-day trial today. No credit card, no commitment.