AI glossary · Agents
Conversation flow
In one sentence
A conversation flow is the predefined path of an automated conversation: the steps, questions and replies a bot follows depending on what the user answers, usually drawn as a branching diagram.
Also known as: chatbot flow · dialogue flow · conversation tree · conversational flow · bot script
What is a conversation flow?
It’s the script of an automated conversation. Before today’s language models, every chatbot worked this way: someone drew a decision tree with every question, every possible answer and where each one leads. The bot doesn’t improvise; it follows the path.
Put it to work
An AI agent on your WhatsApp, CRM included
From €49/mo + VAT
It replies to customers on WhatsApp and logs every conversation in the CRM.
See the WhatsApp agentThe term now sits alongside “agent”, and the two are worth telling apart:
| Conversation flow | AI agent | |
|---|---|---|
| Who decides the next step | The designer, in advance | The model, at each turn |
| If the user goes off script | Gets lost or returns to the menu | Understands and adapts |
| Predictability | Total | High if well scoped |
| Maintenance | Grows with every new case | Adjust instructions and data |
An AI agent doesn’t need a tree for every case: it gets instructions and decides. That doesn’t make flows useless, though.
Related to this term
UTIO WhatsApp
- 3 users and 500 AI conversations
- CRM and shared inbox included
- Built from your website in a minute
- 14-day free trial, no card needed
From €49/mo + VAT
See the WhatsApp agentHow it works
A flow is built from a handful of elements:
- Trigger: what starts it (a message, a keyword, a detected intent).
- Nodes: each step the bot takes, whether a message, a question or an action.
- Conditions: branches based on the reply (yes/no, buttons, validated data).
- Variables: what’s stored along the way (name, date, service).
- Exits: the happy ending, the error path and the human handoff.
The most common mistake is designing only the ideal path. A good flow plans for the user not replying, replying with something else or asking to speak to someone.
Example
A driving school in Nottingham uses a WhatsApp flow for sign-ups. The flow always asks for the same things in the same order: licence type, name, date of birth and acceptance of the privacy policy. That part isn’t left to improvisation because it has legal requirements. Open questions (“how many lessons does someone who’s never driven usually need?”) are handled by an AI agent, which returns to the flow once the person wants to enrol.
Conversation flows at UTIO
UTIO’s agents aren’t built as decision trees: they work from instructions, your business information and rules about what they’re allowed to do. You can build the agent yourself from your website in a minute. Fixed rules exist where they’re useful, for example deciding when to pass a call to a person: by time of day, by topic or when the caller asks.
In the WhatsApp agent, every conversation has an owner in the shared inbox, and autonomy is gradual: no AI, ask first, draft or automatic.
Frequently asked questions
What's the difference between a conversation flow and an AI agent?
A flow follows a path drawn in advance: if the customer answers A, it goes down branch A. An AI agent decides the next step at each turn based on the conversation and its instructions. A flow is predictable; an agent is flexible.
Do conversation flows still matter with language models?
Yes, for steps that must always run the same way: collecting consent, verifying identity, taking a payment or meeting a legal requirement. The usual approach today is to combine them: a flow for the mandatory parts, AI for the open-ended ones.
What do you use to design a conversation flow?
A diagram of boxes and arrows, on paper or in the visual editor of a bot platform. The key is planning what happens when the user doesn't reply as expected.
Related terms
- Chatbot
A chatbot is a program that holds text conversations with people, on a website or on WhatsApp, to answer questions or guide them through a task. It either follows fixed rules or uses a language model to understand what people write.
- Intent
An intent is the goal behind what a person writes or says to a chatbot or AI agent, such as booking, cancelling or asking for a price. Recognising it lets the system decide what to do with the message.
- AI agent
An AI agent is a system that uses a language model to understand a request, decide which steps to take and carry them out with real tools, such as your calendar or CRM, until the task is done rather than just replying.
- Human handoff
Human handoff is the point at which a chatbot or AI agent passes the conversation to a member of staff, together with the context it has gathered, because the case needs it, a rule requires it or the customer asks for it.
- System prompt
A system prompt is the set of fixed instructions an AI model receives before any conversation starts. It sets the model's role, its tone, what it knows about the business and what it isn't allowed to do.
Updated on