Profiles

Automatically collect structured customer data during conversations: field setup, update modes, analytics, and use in the Persona.

Profiles are customer cards that the AI Employee fills in automatically during the conversation. No manual work is needed: while the employee talks with the customer, it extracts the data you need and writes it into a structured table.

You decide exactly what to collect. Name and language are collected by default. Everything else is fields you configure for your needs.

Persona

Use profile data in the employee’s instructions

AI Employees

Manage employees and their settings

Chats

Conversation history with customers

Dashboard

Summary analytics across all employees

“Profiles” refers to two different places in the interface.

• The Profiles tab in the AI Employee’s settings — this is where you configure which fields to collect and when to fill them in. This is configuration.
• The Profiles section in the platform’s left menu — this is where filled-in customer cards are stored. This is an analytics tool.

This page covers both.

---

Video guide

The Profiles feature

Analytics and profiles

A short video: how to configure profile fields in an agent, what’s stored in the Profiles section, and how to use the collected data for analytics.

---

Configuring profile fields

Open the AI Employee → Profiles tab. On this tab, you describe which fields of the customer profile the employee should fill in and when to do it. Fields are tied to a specific agent: different employees can collect different sets of data.

Each field you define here appears as a separate column in the shared Profiles section and is automatically attached to the metadata of every chat for this employee.

Update mode

The first thing to choose is when the employee updates the profile.

Mode	When it updates	When to use it
Realtime	After every customer message	When the employee needs to use the data right away — for example, to address the customer by name or account for preferences in the same conversation
Scheduled	On a schedule (for example, once every 24 hours)	When you need the data only for analytics and reporting

When you choose scheduled mode, additional settings appear: interval, start hour, and time zone. You’ll also see a token usage forecast — how much will be spent per day and per month.

If you’re not sure what to choose, use scheduled updates once every 24 hours. That’s enough for most scenarios and saves a lot of budget compared to Realtime.

---

Field types

Each field has a type that determines whether the employee can update its value.

Type	Behavior	Example
Variable	Updates with every new mention	Language, preferences, status, current interest
Constant	Set once and never changes	Contract number, date of birth, customer type

A constant can’t be changed after it’s first filled in. If you’re not sure, choose Variable.

---

Field parameters

Each field has four parameters:

Parameter	What to enter
Name	A clear field name: “Name,” “Language,” “Request status”
Description	An instruction for the AI — when and what to fill in. This is the most important parameter
Key	Generated automatically; no need to change it
Type	Variable or Constant

The field description is what matters most. Write it from the customer’s perspective: not “when the customer states their language,” but “the language I communicate in.” The more precise the description, the better the employee understands when to fill in the field.

Also specify the answer format. For example: “yes/no” or “choose one of: Russian, English, other.” Without this, the employee may record arbitrary text that’s hard to analyze.

---

How to add a field

1. Open the Profiles tab

   Go to the AI Employee’s card and select the Profiles tab.

2. Add a new field

   Click the button to add a new field.

3. Enter a name

   Give it a clear name — for example, “Name,” “Language,” “Budget.”

4. Fill in the description

   Describe in detail, from the customer’s perspective, when and what to fill in. Specify the answer format.

5. Choose a type

   Variable — if the value can change. Constant — if it’s set once.

6. Save

   Repeat for each field you need. There’s no limit on the number of fields.

Add only the fields you really need — this improves filling accuracy and saves tokens.

---

Field examples

See example field setups

Field	Type	Description for the AI
Name	Variable	My name
Language	Variable	The language I communicate in. Choose one: Russian, English, other
Preferences	Variable	Preferences or wishes I’ve mentioned. Update with every mention
Complaint	Variable	When I talk about any problem. Describe briefly, no more than four words. Always update
Status	Variable	The current request status. Choose one: new, in progress, resolved
Budget	Variable	When I talk about a budget or amount. Record a number
Contract number	Constant	The contract or booking number I gave
Summary	Variable	Two sentences about what the conversation is about. Look at the whole history. Always update

These are examples — the set of fields depends entirely on your business and what matters to track.

---

Profile fields and chat metadata

Fields configured on the agent’s Profiles tab automatically go into the metadata of every chat for this employee. This means:

• In the Profiles section (left menu), you see these fields as table columns.
• In the Persona, you can reference these fields as variables — the employee will insert the values right into the conversation (see Persona).
• In chats, the metadata is available via the chat actions menu → “Customer profile.”

Field types

Field type in settings	Stored in metadata as	Example
Text (Variable)	Free text	Name, City, Summary
Text with answer format (Variable)	One of the defined options	Language: Russian / English / other, Status: new / in progress / resolved
Text (Constant)	A fixed value that doesn’t change after it’s first filled in	Contract number, Date of birth
Number (Variable)	A numeric value	Budget, Number of guests
Short description (Variable)	A few words or one sentence	Complaint (≤ 4 words), Summary (2 sentences)

The field type isn’t set with an explicit selector but through how you word the description for the AI. In the description, specify the answer format — “choose one of: …,” “record a number,” “no more than four words” — and the employee will stick to that format when writing to the metadata.

---

Profile analytics

The Profiles section in the platform’s left menu is an analytics tool. It stores the filled-in cards of all customers across all employees.

KPI tiles

At the top of the page are four summary cards:

Metric	What it shows
Total profiles	The total number of customers for all time
Completeness	The average percentage of filled fields — how fully the employee extracts data
This week	New profiles this week
This month	New profiles this month

View tabs

Above the table are tabs:

• All data — the standard view with all fields for all employees.
• Saved views — custom views with preconfigured filters, columns, and sorting. Handy when you have several analysis scenarios: “Hotel leads,” “Complaints this week,” “Requests in progress.”

Create your own view: set up filters and column order for the task at hand, then save it — next time you’ll open it in one click.

Profiles table

Each row is one customer. Standard columns: profile, employee, Chat ID, name, language, summary. Plus all the custom fields you configured on the Profiles tab of each agent (for example, business industry, purpose of purchasing the service).

Table features:

• Filter by employee — view data for a single employee or branch
• Search by name, field contents, or any text
• Go to chat — click the Chat ID to open the full conversation history
• Column management — hide columns you don’t need
• Drag-and-drop column order (new in v2.1.0) — drag column headers with your mouse to arrange them the way that works for you. The new order is saved in the current view
• Display modes — switch between Cards and Table views

If a field isn’t being filled in, check its description in the employee’s settings. Most likely, the AI doesn’t understand when exactly to record the data.

---

Using profiles in the Persona

You can use profile data right in the employee’s instructions — through profile variables in the Persona section. For example, if the customer’s name is known, the employee can address them personally during the conversation.

This lets you create truly personalized communication: account for preferences, adapt the style, and offer relevant solutions based on the collected data.

---

FAQ

Is Realtime more expensive than scheduled?

Yes, significantly. Realtime updates the profile after every message and uses noticeably more tokens. Scheduled mode processes data in batches, which is more economical.

Can I change a constant's value?

No. That’s the whole point of the Constant type — the value is set once and doesn’t change. If you’re not sure, choose Variable.

Are profiles shared across all employees?

No, each employee keeps its own customer profiles.

How many fields can I add?

There’s no limit, but add only the fields you really need — this improves accuracy and reduces token usage.