Skip to content

Work and systems

For wholesalers and distributorsB2B orders

The e-mail order is transferred to the system without being rewritten.

Orders from customers come in e-mails, PDFs and spreadsheets. The system reads the items itself and checks them in the catalog. The merchant only looks at items that the catalog has not confirmed.

Tested on sample data, without customer data.

test cases with a well-recognized product, unit or substitute
21/21
uncertain positions recorded without human decision
0
for reading one order by AI
€0.0027

Syntalith test results on prepared orders. Last checked: October 22, 2026.

The salesperson transcribed each order item by item.

The customer writes "10 pieces of the last one" or sends his own sheet. Someone has to find the right product in the catalog, check the unit and transfer everything to the system. The error occurs only during shipping.

  1. 01

    The right product

    The customer name is different from the name in the directory.

  2. 02

    Unit

    Pieces, boxes and pallets must match.

  3. 03

    Replacement

    When a product is out of stock, you need to choose an appropriate substitute.

From e-mail to ready order in four steps.

  1. 01

    The mail goes into the queue

    The system reads the content of the email, PDF and spreadsheet.

  2. 02

    Checks the directory

    Each item gets a product, a unit and a possible replacement.

  3. 03

    Stops doubtful

    The uncertain position shows the trader with possible matches.

  4. 04

    Your order is ready

    The system receives a document in which each item is resolved.

What does the system do and what is left to the salesperson.

Makes a system

  • Reads items from email, PDF and spreadsheet
  • Matches products and units from the catalog
  • Prepares a draft of the order

The salesperson decides

  • Items that the catalog has not confirmed
  • Choosing a replacement when there are several
  • Approval of the entire order

Statistics: how much does it cost to manually rewrite orders in 2026.

According to the Syntalith test (2026), the system correctly recognized the product, unit or substitute in 21 out of 21 cases.

Reading one order by AI cost €0.0027 in the test.

No uncertain item entered the system without a human decision.

According to the Central Statistical Office, the average salary in the enterprise sector in September 2026 was PLN 9,266.48 gross. With employer contributions and 168 hours a month, an hour of work costs the company PLN 66.45.

Scenario: 480 orders per month

By hand
480 × 11 min = 88 h of work, PLN 5,848
Reading by AI
480 × 0.0027 PLN = 1.3€0
He stays with people
only items that the catalog has not confirmed

Estimate. Assumptions: 480 orders per month and 11 minutes of manual rewriting per order are sample numbers. We measure exception checking time after launch at the client.

What stays with the team.

A precarious position awaits
Without confirmation in the directory, nothing reaches the system.
The reason for the position
The trader sees why the system has stopped.
Decision history
Every decision has a record: who, when and why.

This is what a salesperson's job looks like.

Screens from the test on sample data.

  1. Otwórz zamówienie z maila
    01Otwórz zamówienie z mailaOperator widzi tylko pozycje do sprawdzenia.
  2. Rozstrzygnij pozycję
    02Rozstrzygnij pozycjęKażda decyzja ma uzasadnienie.
  3. Pobierz gotowe zamówienie
    03Pobierz gotowe zamówienieDo systemu trafia sprawdzony dokument.

Does your team also copy orders from emails?

Describe your process to us. We will show the prototype on your process and your data. The analysis and prototype cost €0, and you know the fixed implementation price before the contract.

What our OpenAI and Anthropic partnerships mean for you.

OpenAI makes ChatGPT and Anthropic makes Claude. We belong to the partnership programs of both companies, so we work with their tools every day and learn from the creators themselves.

OpenAI Select PartnerClaude Partner Network
01We will select a tool for the task
We implement both ChatGPT and Claude. We check both on your documents, and you choose the one that makes fewer corrections and saves more time.
02Knowledge straight from the creators
We learn from official OpenAI and Anthropic courses and materials. Your team learns about the features and settings that work today.
03Licenses without our margin
When we roll out ChatGPT, Claude or Copilot, you buy the licences directly from OpenAI, Anthropic or Microsoft and they invoice you. We add nothing to the licence.
04Certified engineers
On the team, we have an AWS certification in machine learning and an Anthropic certification in systems architecture on Claude.

Participation in programs does not mean that OpenAI or Anthropic endorses our services.