Telegram support automation for automotive logistics
About the company W8 Shipping
W8 Shipping is an international operator for vehicle and cargo delivery from the United States. The company works through six own warehouse and loading complexes near key US ports and, according to its official website, exports more than 120,000 vehicles per year. At this scale, support must quickly preserve context for every car, customer and delivery stage.
Challenge: support across 800 Telegram groups
Build and evolve a large-scale support system as the main customer interaction tool across all stages:
- Centralize questions from around 800 Telegram groups
- Route tickets by topic, department and status
- Speed up answer search across hundreds of thousands of requests
- Automate VIN-code handling and vehicle data workflows
- Give managers detailed analytics on support quality
Solution: from Telegram group to controlled ticket
We built a system where the customer continues communicating in a familiar Telegram group, while the team works in a centralized web workspace. Every question becomes a ticket with a topic, department, status, history, reminder and overdue control. Elasticsearch finds similar requests across hundreds of thousands of records, VIN codes are recognized automatically, and managers see workload, response speed and customer ratings. The work exceeded 500 management hours and 1,500 development hours.
Support system architecture
Chatbot type:
Platforms:
Telegram and centralized tickets
The bot works as an administrator in around 800 Telegram groups. It adds topic menus and service buttons that help customers formulate a request, while the team gets unified routing rules across hundreds of separate chats.
Every question becomes a separate ticket with topic, responsible department and current status. Color coding, filters, reminders and overdue control help managers define priority and avoid losing postponed requests.
Elasticsearch indexes hundreds of thousands of requests and returns relevant answers by keywords. Managers reuse the accumulated team knowledge base instead of preparing every answer from scratch.
A VIN code is automatically extracted from a customer message and becomes a search parameter. If there are several codes, the manager switches between them. Integration with an internal parser also returns ready explanations for known auction-lot notes.
Search, calls and analytics
The web system became the workplace for support managers and team leads. It unites tickets, templates, reminders, calls, work schedules, reporting and customer ratings without switching between disconnected services.
Color-coded statuses, filters, reminders and overdue control show which requests require attention first. Templates are connected to departments, so managers see relevant drafts for a specific topic.
Binotel starts a call directly from the request card. The system stores duration, manager and a link to the call recording, simplifying quality control and dispute review.
Analytics track the number and speed of ticket closures, first response time, message volume and calls. Weekly and monthly reports are generated automatically by predefined parameters.
Work schedules and time zones are included in average performance calculations. A multilingual model lets the team add new languages and centrally edit all texts and buttons for new markets.
Support system functionality
1
Single request pool from around 800 Telegram groups
2
Hundreds of thousands of tickets with topics, departments, statuses and search
3
Elasticsearch for answer search across a large data set
4
Automatic VIN recognition and vehicle-data workflows
5
Answer templates, reminders and overdue ticket control
6
Binotel IP telephony integration and call recordings
7
Multilingual content, manager analytics and automatic reports
