Skip to content

Blog · Published · Updated

AI agents for investigating invoice discrepancies

Short answer

For an invoice discrepancy, an AI agent sets the invoice, the purchase order, the goods receipt and the supplier's emails side by side, points to the document that may explain the difference and writes the accounts payable employee a specific question for the warehouse. It changes no quantities in the system. The agent costs from €3,000 net at Syntalith and, with 400 invoice discrepancies a month at 15 minutes each, pays back in 1.2 months.

How we workDetails ↓

  1. 01Call30 minutes with an engineer, free
  2. 02PrototypeA working system on your process and data, free
  3. 03ImplementationTarget and fixed price in the contract; 50% at signing
  4. 04Acceptance50% once the agreed target is met
net per agent for one process, build paid once
from €3,000
from signing the contract to acceptance
6–16 weeks
free prototype on your cases
€0

We belong to the OpenAI and Anthropic partner programmes and learn through their official courses. What this means ↓

Author: Syntalith · 3 min read

Example: the morning request inbox
Agent · ready

Statistics

Invoices and AI: statistics 2026

1.2 mo.

payback of a €3,000 agent with 400 invoice discrepancies a month: €3,490 of work today, €1,059 with checking and the model. Scenario.

Figures checked: 22 October 2026

What happened to the other packs

A company ordered ten packs of goods and received an invoice for ten. The warehouse system shows a goods receipt for six. An email from the supplier holds a document describing a second delivery of four packs. The accounts payable employee sees a difference of four units, but without that email asks the warehouse about the whole delivery, waits for an answer and only then looks further.

The agent finds the supplier's document and writes a question: „The invoice and the purchase order cover ten packs, while the system shows a goods receipt for six. The supplier's email holds a document for a second delivery of four packs. Did the warehouse receive that delivery?”. It attaches links to the receipt and the email, so the warehouse operative checks a specific batch. The supplier's document helps ask the question, and it does not prove that the warehouse received four packs. When the warehouse operative replies, the agent attaches the explanation to the case, and the next person does not start with the same question.

From a difference to an explanation
  1. 01Differencethe invoice, the purchase order and the goods receipt side by side
  2. 02Documentsthe supplier's emails and attachments about this line
  3. 03Questiona specific request to the warehouse with links
  4. 04Replythe warehouse's explanation on the case
  5. 05Decisionthe person responsible for the invoice workflow decides what happens next

An email from the same supplier may describe a different delivery. When the agent is not sure that a document concerns the same order line, it shows it as a possible lead.

What it costs and when it pays back

Scenario: 400 invoice discrepancies a month
Today
400 invoice discrepancies × 15 min = 100 h (€3,490)
With the system, a person checks the output in 30% of the time
€1,047
Model, 2 calls per case, Claude Sonnet 5.5
€12
Cost of one case
€8.73 today, €2.65 with the system
Saving
€2,431 a month

Payback of €3,000

1.2 months

The volume, the time and the checking share are the scenario's assumptions. €34.90 an hour is the EU average hourly labour cost for 2025 according to Eurostat. Model: 6,000 input and 500 output tokens per call at Anthropic's rates, converted at the ECB rate of 1.1205 USD per euro on 22 October 2026.

The count covers the work around a case: searching, reading, writing and passing things on. What remains is the time to check the output, 30% of today's work in the scenario. Before the quote we measure both times on 20 to 100 cases the team has already handled, in a €0 test that takes 5 working days.

When your current system is enough

An accounting or warehouse system often already matches the invoice with the purchase order and the goods receipt and flags a mismatch. If the receipt is in the system and the employee easily finds it and links it to the invoice, your current workflow is enough. When the problem comes from a misread invoice, fix the reading. An agent makes sense when the explanation regularly takes you outside the system: finding an attachment in the inbox, a conversation about the second delivery and the warehouse's reply.

What to choose for which discrepancies
  • The system matches invoice and receipt and shows the documents

    your current workflow

    no build

  • Errors come from misread invoices

    better invoice reading

    automation

  • Explanations need email and questions to the warehouse

    an agent that gathers documents and questions

    from €3,000

Where to start

The first scope covers quantity discrepancies in one purchasing workflow. On a finished case we check whether the agent found the same document the employee found by hand. We build this as part of AI agents. For the free process review, a description of one case from recent weeks and the tools it ran through is enough. You do not need to send confidential documents, and you get a written plan within 2 working days.

Worked example · Scenario

How we count the payback: cost invoices

We run the same count for your invoices in the free process review.

€1,745

a month for coding 1,000 cost invoices by hand

Payback: 2 mo.

3 minutes per invoice and checking in 15% of the old time are our assumptions. €34.90 an hour is the EU average hourly labour cost for 2025 according to Eurostat, for companies with 10 or more employees. We measured the model cost of €0.004 per invoice in our invoice coding system, converted at the ECB rate of 1.1205 USD per euro on 22 October 2026.

See the full calculation →

This is how we implement it in companies

All case studies →

Client results.

All projects →
foap.comcustomer service

foap.com customer service: 95% of replies are sent automatically.

Users from several countries write to foap.com, in several languages. The system reads the entire conversation and the terms of that person, and then sends a response in that person's language. Questions about money and exceptions go to the team with a ready draft, so they have to check and send them.

the system sends the answers itself, without human typing
95%
time from the arrival of the e-mail to the ready response (median)
30 s
e-mails that require a response receive a ready-made draft
100%
less work for every 1000 answers (estimate)
47 h

Data from foap.com production, October 20-21, 2026. Estimate: 3 minutes for handwritten response.

How foap.com customer service works →
Sample messages. Money matters always come to the person.

Other implementations

jaktako.design

studio projektowe

−87%

Time to look for new customers, every day

4 hours → 30 min

approx. 73 hours less per month (21 days of work)

according to the studio

How the agent loop works →
Signature Estates

real estate

approx. −90%

Operator time spent browsing portals, per day

2–3 hours → 20 min

approx. 45 hours less per month (21 days of work)

respect · contact the seller 5 minutes after the announcement

How the agent finds property listings →
Kuba Koziej, CEO and co-founder of MoreGrowth and board member at Natu.Care

Kuba Koziej

CEO and co-founder, MoreGrowth

„We started with my own AI agent. During the training that followed, I learned to build new applications myself. Today I am bringing this way of working with AI into my business. Working with Syntalith gave me exactly the skills and tools I needed.”

View the original training review on LinkedIn ↗

See the agent on your own cases before you commit.Book a free process review (30 min) →

How we build automation and agents

From a conversation to a working system in four steps.

  1. 01 · 30 min

    Free process review

    €0

    You're talking to an engineer. You receive the written plan within 2 business days.

  2. 02 · 5 days

    Free prototype

    €0

    The system works on 20-100 cases that your team has already handled. You see every result.

  3. 03 · 2–16 weeks

    Build

    50%

    Fixed price and goal written in the contract. First installment upon signing.

  4. 04 · after reaching the goal

    Reception

    50%

    You pay the second installment when the system reaches its target on your cases.

How a free prototype is created

Mon.

We set a goal

What result do you consider good?

Tue–Wed

The system is working

It goes through your affairs, one by one.

Thu

We compare

We check every difference with the team's work.

Fri

Result and price

List of errors and fixed implementation price.

You choose the matters. We anonymize them upon request, they stay in the EU and we delete them after the test.

On the market, such a test usually costs from PLN 3,000 to PLN 8,000.

Book a free process review (30 min)

We establish a prototype during a free analysis.

Data checked on October 25, 2026.

„Is my data safe?”

Yes. You choose where they go: from your company account to the model on your server.

  1. 00Employee's private account

    Customer and company data ends up in an account that the company cannot see or control. There is no data processing agreement, so it is a violation of the GDPR. When an employee leaves, the chat history goes with them.

  2. 01ChatGPT or Claude corporate account

    The provider does not use your data to train models. A copy of the queries remains in his logs for up to 30 days.

  3. 02API with zero retention (ZDR)

    The provider processes the request and does not save anything. It requires his consent, we prepare the application together.

  4. 03Processing in the EU

    Queries are counted and stored in Europe: OpenAI in the EU region, Claude via clouds with the EU region.

  5. 04Model at your place

    The open model runs on your server or in your cloud in the EU. The data does not go outside the company.

What we do with every implementation

  • Confidentiality agreement and data processing agreement before we receive any data.
  • You choose the cases for the prototype and we anonymize them upon request.
  • The data stays in the EU and we delete it after the test.
  • Corporate accounts with an administrator on your side.

When data cannot leave the company

The open model runs on your server or in your cloud in the EU. To choose from:

White-tailed eagle
Polish model, SpeakLeash and ACK Cyfronet AGH
PLLUM
Polish model, HIVE consortium commissioned by the Ministry of Digitization
Qwen
Alibaba
GLM
Zhipu AI
DeepSeek
DeepSeek

We select a model for the task and check it on your cases. We run it, tune it and maintain it.

Source: Bielik and PLLuM, ACK Cyfronet AGH ↗ · PLLuM, Ministry of Digitization ↗

Do you have a lot of your own data? We will fine-tune the small model.

  1. 1 Your examples
  2. 2 tuning
  3. 3 test on new cases
  4. 4 model at your place

When you have thousands of similar emails, documents or reports, we teach a small open model based on your examples. It works for you, responds quickly and costs less to maintain than a large model. We measure quality on issues that the model has not seen while learning.

Przemysław and Piotr did it in large programming and consulting companies. Artem did it before and does it today at Syntalith. It's about models on your own infrastructure, tuning and quality measurement.

You choose the protection level during the interview. We sign the data processing agreement before we see any data.Schedule a free analysis →

AI agents

See the agent on your own cases before you commit.

The agent takes the case through the next steps in your systems, and the actions take effect until the consent of the indicated person is obtained. A prototype for 20-100 cases that your team has already handled costs €0. We write down the purpose and test set in the contract.

See AI agents →
net per agent for one process, build paid once
from €3,000
from signing the contract to acceptance
6–16 weeks
free prototype on your cases
€0

Questions

  • How does an AI agent help explain an invoice discrepancy?

    It sets the invoice, the purchase order, the goods receipt and the supplier's emails side by side, points to the document that may explain the difference and writes a specific question for the warehouse.

  • Does the agent correct quantities in the system?

    No. The agent gathers documents and questions. The person responsible for the invoice workflow changes the goods receipt or decides on the invoice.

  • How much does such an agent cost?

    At Syntalith from €3,000 net. With 400 invoice discrepancies a month at 15 minutes each, the payback is 1.2 months. We agree integrations and maintenance before the quote, and the model costs €12 a month.

  • When is your current system enough?

    When it matches the invoice with the goods receipt and shows all the documents. An agent makes sense when the explanation needs email and conversations with the warehouse.

Let's talk

Just write, call or choose a date.

We work fast because we hate corporate cotton wool. We don't have salespeople or managers to keep an eye on managers. You talk immediately to engineers who then build your system. We respect each other's time and yours too.

We will accept 3 more companies by the end of the year.

Mon–Fri 9:00–17:00 CET

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.