Conversational Order Assistant
Watch the videoOverview
A conversational chat assistant that takes a return in whatever order the customer explains it. An LLM Agent asks for whatever is missing, and an ordinary automation wired to its tools port looks the order up mid conversation and hands the answer back.
Conversational Order Assistant
Conversational Order Assistant takes a product return the way a person would explain it: in whatever order it comes out. Nobody draws the questions. An LLM Agent is told what to collect, and it asks for whatever is missing.
The part worth copying is the branch on the agent's tools port. A Tool In node declares a tool with a name, a description and a JSON schema; behind it sits an ordinary automation; a Tool Out node hands the result back. The model can call it mid sentence, which is how an assistant looks something up in your systems while it is still talking to the customer.
What Conversational Order Assistant can do
- Hold a free-form conversation through the Chat Assistant UI, with no fixed question order
- Work out what is still missing and ask for it, in plain language
- Call your automation as a tool, and use what comes back in its next reply
- Run on Robomotion AI Credits out of the box, or on your own API key from the Vault
- Keep every turn in one conversation through the session id
Behind the scenes
Chat In hands the message to the LLM Agent, which gets the text as its user prompt and the session id so the turns are one conversation. Its instructions say what to collect and how to behave, and nothing else in the flow decides the order of the questions.
An LLM Agent has three outputs before its ordinary one: sub-agents, tools and callbacks. Anything wired to tools becomes something the model can decide to call. Here that is Look Up An Order, whose JSON schema tells the model it takes an eight digit order number, followed by Find The Order and Hand It Back.
The ordinary output goes to a Text node that says the reply, and Chat Out ends the session.
Setup Guide
- Configure Credentials: This template has Use Robomotion AI Credits switched on, so it runs with no key. To bring your own, turn that off and point the LLM Agent's API Key at a Vault item.
- Version & Publish: Create a new version of this flow and Publish it, because an agent can only point at a published version.
- Create Agent: Go to Admin Console > Agents > Create Agent. Pick this flow and its version, choose Conversational mode, and add a few sample questions — they become the opening prompts on the chat page.
- Install Desktop App: Download the Robomotion Desktop App from robomotion.io/downloads and log in to your workspace.
- Connect: Refresh the robot list, find the Application Robot the agent created, and connect it.
- Run: Press Play here in the Designer so you can watch the nodes light up, or start it from the agent card.
- Launch Chat: Open the Agents screen, find your agent, and press Open.
Try it
Say something like "the boots I ordered turned up with a broken zip" and let it ask you for the rest. The order numbers this template knows are 48120677 and 48120691.
Customization
Find The Order is the node to replace first. It holds two orders in a literal so the template runs with no setup; swap it for a portal login, a database query or an API call and change nothing else, as long as it still sets msg.result.
To give the model more to work with, add another Tool In → your automation → Tool Out branch off the same tools port. Each Tool In needs its own name, description and JSON schema — that description is the only thing the model reads when it decides whether to call it, so it is worth writing carefully.
Double-click the Order Assistant node to edit its instructions, or change Model Name to any of the curated models.
More ai templates
See all 8 →- Guided Returns DeskA guided chat assistant that opens a product return, one question at a time. Uses Chat Assistant widget nodes for the questions, checks the order number while the customer is still in the conversation, and loops back to the question when it does not recognise one instead of throwing an exception.
- Knowledge Base EndpointAsk your own documents a question over HTTP. A Robomotion Knowledge Base is indexed by a workspace agent on your machine and searched locally by the robot; one Query Knowledge Base node returns the passages that answer a question, each with the document and heading it came from. Http In and Http Out turn the flow into a service anything can call, and a Min Score fence means it answers "I could not find that" instead of inventing something. Ships four sample documents and a thirty-line Python client.
- Tax Portal MCP ServerServes an ordinary RPA automation to an AI assistant as MCP tools, so a model can ask a question of a system it could never sign into. One Listen HTTP node makes the flow a stateless MCP server; three Tool In nodes declare the contracts a model reads, with no arguments, one required argument and one optional. Behind them a sign in subflow answers a government portal's tax number, password and one-time code from the vault, and two more subflows read VAT returns, penalties and electronic invoices off the page. The caller is given a bearer token and nothing else, and Tool Out's Is Error flag lets the flow refuse a company it holds no mandate for instead of returning an empty answer the model will fill in for itself.
- RAG with DeepSeekRetrieval-augmented generation over your own documents, with the retrieval handed to the agent instead of done for it. Documents are read, chunked, embedded and stored in LanceDB - an embedded vector database that is just a directory on disk. A DeepSeek Agent then answers questions through a search_knowledge tool built from ordinary flow nodes, writing its own queries and searching as many times as a question needs, with the source document named against every fact.
- Self-Learning Invoice AgentAn invoice-processing DeepSeek Agent that teaches itself vendors. The first invoice from a supplier is worked out from scratch and the procedure is written down as a learned skill in the agent's memory; every invoice after runs on what it learned. Extracted fields are recorded to a CSV ledger through a Tool In / Tool Out pair, and memory syncs to the workspace so a whole fleet can run what one robot learned.
- Calorie Coach AgentA personal nutrition tracking assistant powered by Chat Assistant and LLM Agent nodes. Uses a SQLite database to log meals, manage user profiles, and track daily calorie and macro goals through a natural conversational interface with support for photo-based food logging.