Telegram Bot for a B2B Company
Introduction
This project is a multifunctional Telegram bot developed for a B2B company. The main goal was to automate and simplify interactions with clients and partners by providing them with a convenient tool for obtaining information and solving common tasks. The bot serves as the primary point of contact, reducing the workload on managers and speeding up processes.
As the project is under an NDA, the portfolio will only feature a video demonstrating its functionality.
Links
- Git Repository: Private (NDA)
Technology Stack Selection
NestJS with TypeScript was chosen for development. This decision ensured strong typing, a clear modular architecture, and scalability, which is crucial for the project's maintenance and future development.
For interacting with the Telegram API, the Telegraf.js library was used with the nestjs-telegraf integration. It provides a user-friendly interface for all Telegram features, including interactive scenes/wizards.
For deployment, I set up containerization using Docker and Docker Compose, as well as a CI/CD pipeline in Drone CI. This allowed for the automation of building and deploying the application with each update.
Modules and Functionality
The bot's architecture is built on modules, each responsible for its own logical part.
| Module | Description |
|---|---|
| Start | Handles welcoming new users and basic navigation. |
| Info | Provides general information about the company, services, and partners. |
| FAQ | Implements a section with frequently asked questions for user self-service. |
| Support | Contains an interactive wizard for creating technical support tickets. |
| Khabarovsk | A region-specific module with a wizard for submitting verification requests. |
| Admin | Functionality for bot administrators (e.g., broadcasts or user management). |
Workflow
During the project, I completed the following tasks:
- Designed and implemented the modular application architecture using NestJS.
- Developed interactive multi-step wizards to collect data from users, such as for verification requests or support tickets.
- Implemented an access control system using guards to separate logic for administrators and regular users.
- Configured dialogue state management (session management) to ensure the correct operation of the wizards.
- Prepared and configured the deployment environment by writing a
Dockerfileanddocker-compose.yml. - Set up CI/CD in Drone CI to automate the building and deployment of the project to the server.