Telegram Lead generation bot

General Logic

The scheme represents a lead qualification funnel composed of two sequential AI dialogues:

1. The bot greets the user and initiates the collection of information about their task.
2. The first AI agent step-by-step identifies the need, project scope, deadlines, and budget, as well as determines the client’s readiness level.
3. After qualification is completed, the data is transferred to the second AI agent.
4. The second agent uses the collected parameters and knowledge base to select suitable services or plans, present a personalized offer, and determine the client’s next step.
5. After completing the second dialogue, the lead is handed over to an online manager.
6. The user receives a message confirming their request is being processed.
7. The next step is intended to send data to an external service via webhook; however, this request is currently disabled and requires configuration.

The scheme also allows for further CRM integration or automation via webhook, for example, to create a contact and deal.

Lead Qualification Group

This part is initiated from the bot’s starting point and performs the role of primary requirements gathering.

The AI agent must ask the user four questions strictly in sequence:

1. What task or problem needs to be solved.
2. What is the approximate scale of the project.
3. How urgent is the solution required.
4. What budget is expected.

Each question has short answer options presented as buttons. The agent should ask only one question at a time, not repeat already obtained answers, and account for situations where the user provides several parameters in one message.

Collected data is saved in a structured format:

• direction or type of task;
• project scale;
• implementation timeline;
• budget;
• lead status: hot, warm, or cold.

If the user refuses to answer a specific question, the agent should first try to clarify the information; after a second refusal, the value should be saved as undefined and the dialogue continued. The user can modify previously provided parameters if needed.

After all fields are filled, the bot thanks the user and passes the results to the next AI agent. The history of this dialogue is stored for up to 48 hours, and the maximum dialogue length is 50 messages.

Inline and reply buttons are configured for answers in this part. Free text input is disabled at the block settings level, so to ensure full functionality, verify how the channel handles user text responses and whether the available buttons meet the scenario requirements.

Solution Selection Group

The second part receives the results of the primary qualification and uses them to consult the user.

The context passed to the second agent includes:

• client’s task;
• project scale or volume;
• desired timeline;
• budget;
• lead readiness status.

Based on this data, the agent should:

1. Briefly demonstrate understanding of the client’s need.
2. Offer one or two most suitable options.
3. Describe their advantages considering the project scale.
4. Propose the next action via CTA buttons.

Possible user actions:

• select a suitable solution;
• request a demonstration or call;
• request a quote or commercial proposal;
• contact a manager;
• request an individual consultation.

The outcome of this dialogue includes:

• selected tariff, package, or individual solution;
• client’s name;
• phone or Telegram contact;
• client’s intention: ready to purchase, wants a demonstration, or needs an individual consultation;
• additional requirements, constraints, and agreements.

The agent should not invent missing services, prices, guarantees, or conditions. If there is no exact match in the knowledge base, it should offer an approximate range or custom calculation.

Before ending the dialogue, the agent must collect contact details. If the client refuses, available information and the refusal reason are saved in additional notes.

Manager Transfer Group

After consultation completion, the dialogue transfer block to the employee is triggered.

Transfer settings:

• request is sent to any available online manager;
• distribution is based on the lowest current workload;
• automatic replies are disabled;
• dialogue rating is not requested;
• the manager receives a separate notification with key lead information.

The notification includes:

• user’s name and their given name if collected;
• phone or Telegram contact;
• initial task;
• budget;
• selected solution;
• client’s readiness level.

The user receives a message stating the manager is already reviewing the details and will contact them within 15 minutes.

For this part to work, managers must be configured in the system, have chat access, and be able to receive requests online. If leads need to be directed to a specific employee, the current distribution mode among all available managers should be changed.

Webhook and External Data Transfer Group

After transferring to the manager, the scheme contains a separate API request block. Transition to it occurs after any message in the previous group.

A POST request is configured to a placeholder address:

https://your_website.com

The request body is expected to include:

• user name;
• client name;
• contact data;
• task type;
• budget;
• selected solution;
• client intention.

Currently, the API block is disabled, so data is not actually sent to the external service.

The comment suggests using webhooks for integration with CRM or automation systems such as HubSpot, Salesforce, Pipedrive, n8n, or Make. This allows automatic creation of contacts, deals, and attaching qualification results to them.

AI Dialogue and Knowledge Base

The scheme uses two AI dialogues. The second agent operates with the knowledge base enabled, so to ensure full bot functionality, the knowledge base must be populated with up-to-date materials.

In the bot settings on the knowledge base page, you should upload service catalogs, price lists, tariff descriptions, case studies, documentation, and other materials on which the proposals are based. Also, ensure the knowledge base usage is enabled in the corresponding AI dialogue settings.

The knowledge base should be regularly updated to keep proposals, prices, and terms current.

Integrations and Required Settings

The following external and system capabilities are involved:

• OpenAI AI service for both dialogues;
• platform knowledge base for solution and term selection;
• chat transfer to online manager;
• potential webhook to send the lead to an external system.

For AI operation, a configured OpenAI connection with a valid API key and access to the chosen model is required. Depending on the provider’s conditions, balance top-up or active billing may be needed.

For manager transfer, operators and their availability must be set up in the workspace.

To enable CRM or automation integration, you need to:

• replace the test URL with a working webhook address;
• enable the API block;
• add authorization headers or secret keys if required;
• coordinate field format with the CRM or the n8n/Make scenario;
• verify handling of external service responses and delivery errors;
• if using CRM, perform necessary access setups and obtain integration tokens.