Global Legend — who is the bot
The Flow Settings → AI Settings → Global Legend section is the foundation of the AI bot’s personality. This is where you describe who the bot is: what its name is, where it lives, what it does, why it
The Flow Settings → AI Settings → Global Legend section is the foundation of the AI bot’s personality. This is where you describe who the bot is: what its name is, where it lives, what it does, why it communicates with the lead, and what role it plays in the conversation.
Global Legend defines:
- how the bot introduces itself
- what tone it uses to communicate
- how it explains the product
- how it answers personal questions
- how “human” and consistent it appears in a long conversation.
The more realistic and specific the legend is, the more stable the bot’s behavior will be.
An empty or abstract legend such as:
“You are AI assistant”
usually leads to generic, robotic, and contradictory answers.
Critical
Global Legend must always be written only in English — even if the bot communicates with leads in Russian, Hindi, Spanish, or any other language.
AI models follow system instructions much better when they are written in English.
The language used to communicate with the lead is set separately in Response Style — How the Bot Speaks.
For example:
Always answer in Hindi or English.
Screenshot: the Flow Settings → AI Settings section, the Global Legend tab with the text field filled in (demo data).
Why Global Legend is needed
Without it, the model:
- does not understand who it is
- starts improvising
- changes its communication style
- gets confused about the backstory
- answers too generally.
A good backstory makes answers:
- consistent
- realistic
- emotionally coherent
- stable even when faced with unexpected questions.
This is especially noticeable when a lead asks:
- “Tell me about yourself”
- “Where are you from?”
- “Why are you doing this?”
- “How did you start?”
If the backstory is written well, the AI responds confidently and naturally.
What a good backstory should include
Basic structure:
Element | Why it is needed |
Name and age | Establish the identity |
City and country | Create cultural context |
Profession / role | Define the communication tone |
Background | Makes the character realistic |
Transition to current activity | Explains the motivation |
Current activity | Shows how the bot helps |
Target audience | Helps AI understand who it is talking to |
Mission | Shapes the emotional tone |
gicon:Lightbulb
Tip
Add only the details that actually affect the dialogue. Useful: country, profession, backstory, mission, experience, audience type. Usually useless: eye color, favorite food, random facts unrelated to the product. AI works better with a realistic backstory like “worked at a bank, burned out, moved into trading” than with a generic “expert with 10 years of experience.”
Real example from the product
This example is used in Demo Space Chatterfy.
Rahul Malhotra is 30 years old from Pune, India. He previously worked as an outsourcing engineer on Western projects, a good job and paid well without rush growth in freelance markets. He always wanted to upgrade his skill in coding to get better salary but never finished a single course completely. Through some difficult times when his salary at the agency was downgraded, he found PocketOption — a fast and profitable trading platform with $1 minimum trade. Now Rahul shares his experience and signal-based trading strategy in his Private Signals Group on Telegram, helping other Indian users start their journey on PocketOption with minimum risk.
Why this backstory works
What is done right here:
Element | Why it matters |
Specific name and city | AI gets a realistic anchor |
Previous job | Gives the character depth |
Transition story | Makes the current role logical |
Product specified | AI understands the offer context |
There is an audience | Indian users |
There is a mission | “helping with minimum risk” |
As a result, the bot:
- responds consistently
- talks about the product more confidently
- handles small talk more naturally
- is less likely to “break” on unexpected questions.
Short version of the backstory
Sometimes a compact version is enough.
Your name is Abu Ali — an Indian trader from Hyderabad. Your mission is to teach beginners how to start trading safely, register, make their first deposit, and learn consistent habits. You focus on building confidence and discipline rather than chasing luck. You run The Rising Traders project — an online mentorship space where Indian traders support each other.
This format works when:
- deep background is not needed
- the role and mission are more important
- the bot works in short conversations.
Example for another offer
Maria Santos is 28 years old from São Paulo, Brazil. She studied marketing but spent five years as a part-time English teacher because of low job market. After trying multiple side incomes she discovered Bet365 sports trading, where her statistical mind finally found a fit. Now Maria runs her Brazil Sports Signals Group on Telegram, helping Brazilian fans turn their passion for football into a stable side income through smart bets.
Examples by segment
E-commerce
Sarah Chen is 34 years old from London, UK. She worked in retail customer service for 8 years before joining ShopX as their senior support manager. She loves helping people find exactly what they need and always remembers regular customers by name. Sarah handles incoming questions about products, orders, and returns in the ShopX Telegram chat.
Sales
Alex Kovalenko is 32 years old from Kyiv, Ukraine. He has 6 years experience selling SaaS products to mid-market companies. He believes in qualifying leads carefully — wasting time on poor-fit clients hurts everyone. Alex is the first contact for new leads of CompanyName, qualifying their needs and budget before passing to senior managers.
Support
Miguel Rodriguez is 29 years old from Madrid, Spain. He studied computer science and worked as a backend developer for 4 years before transitioning to technical support. He loves explaining complex tech in simple terms. Miguel helps users of ProductX troubleshoot issues, set up integrations, and understand the platform.
How to Write a Global Legend
1. Define the character's role
First, answer:
- who the bot is
- what role it plays
- how it should be perceived.
For example:
- mentor
- support manager
- sales consultant
- trader
- community admin.
2. Add background information
2–3 sentences:
- what they did before
- how they got into the current field
- why they started doing it.
This helps AI answer personal questions naturally.
3. Describe current activities
What the bot does now:
- runs a Telegram group
- helps beginners
- sells a product
- provides consultations
- supports customers.
4. Specify the audience and mission
Who the bot helps:
- beginners
- Indian traders
- small business owners
- VIP clients.
And why:
- teach safely
- simplify setup
- help avoid mistakes
- build confidence.
5. Translate into English
Global Legend should be:
- coherent text
- without bullets
- in natural English.
Working structure: name → background → transition → current role → mission.
6. Test
After saving, run 5–10 test conversations. Pay special attention to questions: “Who are you?”, “Why are you doing this?”, “Where do you live?”, “Tell me about yourself.” If the bot answers inconsistently, the legend is too weak or contradictory.
What to avoid
Important
A poor legend makes the bot unstable.
Common mistakes:
Mistake | What happens |
Role is too generic | Responses become formulaic |
Conflict with Response Style | Tone “drifts” |
Several incompatible roles | AI gets confused |
Text is too long | Some instructions are ignored |
Slang without context | AI interprets it incorrectly |
What to keep in mind
Important
If the backstory is written in abstract terms, AI will start inventing details on its own. A simple rule works with models: provide context — the model uses it. Do not provide it — the model improvises. And that improvisation often does not match the image the business needs.
Frequently asked questions
Related
Was this article helpful?