Chemical manufacturing · Analytics and management
How a chemical manufacturer gave directors access to 1C data - without a finance specialist and without waiting
TLDR
A manufacturer of industrial chemicals with about 150 employees. The commercial director and CEO regularly needed operational analytics from 1C: margin by item, inventory, accounts receivable, and plan completion. Every request went to a finance specialist; the answer came in 2-4 hours, sometimes the next day.
BackendFlow developed an AI assistant connected to 1C:Complex Automation. Directors ask questions in a chat in normal language, and the system answers in seconds.
Result: analytics on any data slice in 20 seconds, at any time, without involving a finance specialist.
Context
The company manufactures industrial chemicals and works with corporate customers. The commercial cycle is short: clients request terms, compare with competitors, and make decisions quickly.
All data on sales, inventory, margins, and receivables lives in 1C:Complex Automation. The system covers the entire cycle, from raw material purchasing to finished goods shipment. The data is current and the structure is in place.
The problem was not the data - it was access to it.
To get the required slice, the commercial director went to a finance specialist. The specialist built a report in 1C: 2-4 hours for urgent requests, until the next day for planned ones. By the meeting, the data could already be stale. Decisions about discounts, shipment priorities, and work with debtors were delayed or based on last week's numbers.
"I need to make decisions here and now, not wait until a report is built"
Task
Give the commercial director and CEO direct access to 1C data - without intermediaries, without waiting, and in a convenient format.
Typical requests the system needed to handle:
- Margin by product groups for the current month
- Finished goods inventory by SKU
- Accounts receivable by counterparty with aging breakdown
- Sales plan completion by manager
- Comparison of the current period with the same period last year
Solution
BackendFlow audited the data structure in 1C:Complex Automation - registers, reference books, and the logic for storing analytical data. We fixed the top 30 most frequent management requests, required answer formats, and accuracy requirements.
We developed an AI assistant connected to 1C through an API, without changing the system configuration.
A director writes a free-form request:
"Show the margin for the solvent group for October and compare it with September"
The assistant interprets the request, accesses the required 1C registers, performs the calculation, and returns a structured answer with commentary on the key changes.
The interface is a web chat with a mobile version. It is available 24/7, request history is saved, and the user can clarify or expand a previous answer.
Insights
The most frequent requests turned out to be the simplest.
Most daily management questions are 5-7 standard slices they check regularly. Formalizing them for the system took less time than expected.
The finance specialist stopped being a dispatcher.
Before implementation, a significant part of the finance specialist's time went into building standard reports for executives. After launch, that load was removed, and the specialist focused on analytics that require expert judgment.
Decisions became faster.
The commercial director started checking margin directly during negotiations with a client, without postponing the discount decision until the next day.
Results
| Metric | Before | After |
|---|---|---|
| Time to get analytics | 2-4 hours | 20 seconds |
| Who executes the request | Finance specialist | Assistant |
| Availability | Business hours | 24/7 |
| Data freshness | 1-3 days | Real time |
What is next
The next step is connecting the production loop: data on output, defects, and capacity utilization. This will give the CEO a full operational picture in one tool.