AI analytics on top of 1C in Telegram is a chat where the owner asks a question in plain words and gets data from the accounting system within a few seconds, without waiting for an export from the accountant. We built this setup for AVENUE, a chain of six premium clothing boutiques in Tajikistan. Below: how it works, what came out of it and what you can carry over to your own business.
Key takeaways
AVENUE is a premium clothing boutique chain in Tajikistan: six stores carrying Lacoste, BOSS, Armani, Boggi and other European brands. The project is described in more detail on the AVENUE case page. The business had the data; the problem was reaching it.
All the information sat in 1C, tills, work chats and Excel. To get an answer to a simple question, the owner had to go through the same chain every time.
Sometimes one report took a full day, and evenings and weekends left gaps in that chain. The figures existed, but each time they had to be assembled by hand and passed on through a person.
The owner's request “I just want to ask a question and get an answer right away.” We started from that sentence when we designed the system.
The task can be described with four “nots”. The owner wanted to stop depending on the calendar and on the mood of spreadsheets.
Instead, simply type into Telegram: “Show me August sales by store” — and get the answer, in a clear and simple form, within seconds. The customer did not need yet another report; they needed an answer to a question they had worded themselves.
Several requirements follow from this: answer in a familiar channel, understand free-form wording, require no knowledge of 1C and work when the accounting team is away.
We connected 1C, AI and Telegram. Instead of several disconnected processes there is now a single point of control: the system understands a request in plain human language, works out which data it needs, pulls it from 1C and returns the result with no manual report preparation.
The format of the answer fits the question. For a quick answer, a short figure arrives. For a deeper look, the system sends extended analytics or an Excel file.
The system only reads data and changes nothing in 1C: it is built on top of AVENUE's working 1C setup. For the owner this means a question in the chat cannot damage the books. How AI is connected to 1C, CRM and ERP, which permissions it gets and what it sees is covered in the article on integrating AI with CRM, ERP and 1C.
Here are examples of requests the system answers today.
Look at how the questions are phrased: they are not names of 1C reports but what a person would ask a colleague. Nobody has to learn an interface; asking is enough.
A report answers a question that someone worded in advance: it was set up, named and put in a menu. A question in the chat is worded at the moment it comes up, in the language the person thinks in.
For the owner the difference is that there is no need to know which report to open, which filters to set or which column to look at. “Which store is down compared to last month?” is the whole request. Finding the right data becomes the system's job, not the person's.
The limits matter as much as the capabilities. The system only reads: it does not post documents, correct stock balances or change prices. That is why the owner can be given access without worrying about the database. And trust in the answers is checked before launch: the bot's figures are compared with the accountant's reports.
Reports were only the first step. The next level is operational control: we connected bots to the company's work chats. Information that used to be scattered across several chats is now collected and structured automatically.
The team passes all of this through the system, and the owner gets the data in one place. The bots run on the Telegram bot platform, and how their events reach accounting systems and CRM is described in the article on integrating a Telegram bot with a CRM.
Result Once data began collecting automatically and being checked regularly, the system helped surface errors in operations, discrepancies, weak spots and gaps. Things that used to stay hidden inside a large volume of information became visible almost immediately.
This is how report automation turned into a business control tool: checks run regularly and without manual work, not only when someone decides to open a spreadsheet.
We did not just connect a Telegram bot to 1C. We built the owner a digital assistant for running the business: it works with the company's data around the clock and lets the owner ask a question, get the data, spot a problem and make a decision at any moment.
| What changes | Before | Now |
|---|---|---|
| Access to information | Waiting for an export | Instant access |
| Path to an answer | Dozens of messages | One request |
| Analysis | Manual work through a spreadsheet | Automatic analytics |
| Basis for decisions | Managing by gut feeling | Data-driven management |
The table works well as a list for a conversation with your accounting team: if the left column has more rows than you would like today, those are your automation tasks.
The main change is not speed but habit: a question can be asked the moment it comes up, not when the accounting team finds the time.
The system keeps developing. Next in line are these areas.
The order is clear: first questions about sales and stock, then money, expenses and documents. Based on this case we built the Syntra Control product.
The idea is wider than a retail chain. It is worth considering wherever the figures sit in 1C and the owner gets them through people and files. If you need a regular summary every morning rather than a live conversation, see the article on 1C reports in Telegram for the business owner. Before you start, go through this list.
Tip Start with a list of questions, not with technology. For a week, write down every question you ask staff about figures: you will see right away which ones repeat, and those are the ones to start with.
The order in which AVENUE developed will suit another company too: it goes from simple and safe to more complex.
You can measure the result without complex metrics: how often you still ask staff for exports and how quickly you notice discrepancies.
We start with your 1C database and the owner's list of questions. We already work with 1C:Roznitsa, and other configurations are connected after we review your database, before work starts. Before launch we compare the bot's figures with your accountant's reports, so from day one it is clear which answers can be trusted.
For a typical task there is the Syntra Control product, which grew out of the AVENUE case. For a non-standard task we build AI assistants and Telegram bots to order. Tell us which questions you ask your accountant today, and we will show which of them can be handed to the system.
Let’s discuss your project
Tell us what you need, and we will estimate the timeline and cost and suggest a solution.
We already work with 1C:Roznitsa. Other configurations are connected after we review your database, and that review happens before work starts, so you know in advance what is achievable.
No, the system only reads data and changes nothing in 1C. It is built on top of a working 1C setup, so a question in the chat cannot affect the books.
Before launch we compare the bot's answers with your accountant's reports. Discrepancies are sorted out before the owner starts relying on the answers in daily work.
For a quick question the system sends a short figure, and for a deeper look it sends extended analytics or an Excel file. The format depends on what exactly you asked.
At AVENUE part of the information lived in chats: expenses, till operations, documents and confirmations. The bot collects and structures them automatically, and the team passes data right in the working conversation.
Yes, the digital assistant works around the clock. Evenings and weekends used to mean waiting for answers; now a question can be asked at any moment.
