Updated: 2026-11-13

## Current prices and model costs

[Canonical price list](https://syntalith.ai/en/pricing)

AI apps are operated by your team; automations and agents do the work without it. We decide the form after reviewing the requirements. All prices net.

- ChatGPT or Claude rollout, team of up to 10 people: from €1,150 net, paid once. 2 weeks.

- ChatGPT or Claude rollout, company, 11 to 40 people: from €2,300 net, paid once. 3–4 weeks.

- ChatGPT or Claude rollout, organisation, over 40 people: from €4,600 net, paid once. 6–8 weeks.

- Process review: €0. 30 min + summary within 2 business days.

- First test on your process and data: €0. 5 business days.

- Process audit: from €600. 5 business days.

- AI Automations: from €3,000; for simple, rule-driven processes. 2–6 weeks.

- AI Agents: from €3,000. 6–16 weeks.

- AI Apps: from €3,000; typically €3,000–7,000. 4–10 weeks.

- AI Training: from €1,150/day. 1–2 days.

- AI Act and GDPR audit: from €700. Scope agreed after a review.

- AI-Native course: quoted after a call; paid course · net. 1:1 or company · on your repository.

- Maintenance: priced per project. Optional, agreed during the rollout.

Is maintenance cost, including AI models, predictable?

The cost depends on case volume, models and infrastructure. We estimate it after reviewing the workflow and running trials. Before signing, we agree on billing, usage monitoring and any limits.

Starting prices are not project quotes. Check the current service page and retain currency, unit, net/VAT status and scope qualifications. Older search results may describe a previous offer.

[Assistant directory](https://syntalith.ai/llms.txt)

# Company Brain AI. Shared knowledge for people and agents.

Canonical URL: https://syntalith.ai/en/services/ai-apps/company-brain

We combine documents, decisions and events from Confluence, SharePoint, Google Drive, CRM and ERP into one company memory. Employees and AI agents get answers with sources, within their permissions.

net for the first process, fixed price before the contract: from €3,000

from contract to acceptance on your process and your data: 4–10 weeks

[Book a free process review (30 min)](https://meet.syntalith.ai/syntalith/en?utm_source=syntalith&utm_medium=site&utm_campaign=discovery&utm_content=service-firmowy-mozg-ai-hero)

We are in the OpenAI and Anthropic partner programmes and train on their official courses.

How we work

1. Call · 30 minutes with an engineer, free
2. Prototype · A working system on your process and your data, free
3. Build · Target and fixed price in the contract, 50% at signing
4. Acceptance · 50% once the agreed target is met

## The company's knowledge is in several systems. Every new person starts from scratch.

The Company Brain combines the sources you already have into one memory: facts, decisions, policies and case histories, each record with its source and owner. An employee asks and gets an answer. The agent checks the policy and the case history before taking the next step.

95%

less time to gather information from reports with AI, according to Anthropic's estimate from 100,000 conversations (2025).

[Anthropic, Estimating AI productivity gains (2025) ↗](https://www.anthropic.com/research/estimating-productivity-gains)

## Memory. Context. Action.

1. It remembers what was agreed.
  Facts, decisions and rules with date, source and owner. It distinguishes current from historical records.
2. It gives you context before you act.
  The person or agent gets the status of the case and the applicable rules before taking the next step.
3. First the permissions, then the answer.
  The system checks who is asking before opening the document. Without a basis in the sources, it does not answer.

## Company memory in action.

[All case studies](https://syntalith.ai/en/work)

- Company knowledge with access controls
  750
  question and access-card pairs checked in the test
  An employee asks a question and gets a sourced answer that respects their document access.
  [How it works](https://syntalith.ai/en/work/company-memory-with-permissions)
- Management reporting
  numbers without a source in the test reports
  The controller opens a prepared report and can trace every number back to its source.
  [How it works](https://syntalith.ai/en/work/executive-report-no-vanity-numbers)
- Client document intake
  12
  cases handled across several days of intake
  Every new document updates the same case file, so staff can see what is missing and what needs clarification.
  [How it works](https://syntalith.ai/en/work/client-document-intake)

## Statistics: what does searching for and passing on data cost in 2026?

In the Syntalith test (2026), the app answered 750 questions and never answered from a document the asker could not open. Half of the local model's answers were ready in under 521 ms.

The Syntalith test reports (2026) contained no number without a source.

In the Syntalith test (2026), AI read one request for quotation for PLN 0.0034.

According to Statistics Poland, 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.

Sources: [Statistics Poland, Average employment and salary in the enterprise sector in September 2026. ↗](https://publikacje.new.stat.gov.pl/portal-publikacje/przecietne-zatrudnienie-i-wynagrodzenie-w-sektorze-przedsiebiorstw-we-wrzesniu-2026-r) · [Syntalith case studies](https://syntalith.ai/en/work)

Data checked on October 25, 2026.

750 test questions, each with its own access profile, and no answer outside the asker's permissions

4/4 comments with an unsourced number stopped before the report went out

Scenario: cost of the same work in a month

Inquiries: 160 per month

- By hand: PLN 7,974
- Read by AI: PLN 0.54

By hand: PLN 7,974, Read by AI: PLN 0.54

Monthly closing in controlling

- By hand: PLN 5,316
- AI comment: approx. PLN 0.0010

By hand: PLN 5,316, AI comment: approx. PLN 0.0010

Estimate. The numbers of questions, inquiries and hours are example assumptions taken from our case studies. An hour of work costs PLN 66.45. We price the server for the local model separately.

## How much does a Company Brain cost?

The company's first memory process costs from €3,000 net, and the typical range is €3,000–17,500 net.

Price: from €3,000 net

Implementation usually takes 4–10 weeks. The price depends on the number of sources, roles, agents and activities, and you receive a fixed quote after analysis, before the contract.

When you pay

€0 50% 50% 4 wk 10 wk

Prototype · 5 business days on your cases: €0

Contract · at signing, then 4–10 weeks of build: 50%

Acceptance · once the system meets the target in the contract: 50%

[Full pricing](https://syntalith.ai/en/pricing)

## “Why should I trust you?”

Because the system is built by engineers with more than 22 years of experience who have shipped production systems since 2018.

You talk to the people who then write the code. They have worked at large companies and built systems that run under heavy traffic.

You won't spend money on a prototype that goes nowhere.

First a free call, then a free prototype on your process. If we see no return, we say so in writing and don't build. You pay only for a system that makes sense.

A Polish and Ukrainian team of engineers.

Artem, Przemek and Piotr studied computer science together at Lublin University of Technology from 2017 to 2021. Artem and Serhii founded Syntalith together. Both are from Kyiv and have lived in Poland for years. We work with companies in Europe, the United States and elsewhere, in Polish and English, under GDPR. Contracts are signed with Syntalith sp. z o.o. under Polish law.

The team's careers, year by year

2017–2021: Artem, Przemek and Piotr study computer science together at Lublin University of Technology.

Serhii Ivanchatenko · 7+ years [LinkedIn ↗](https://www.linkedin.com/in/serhii-i-a8b676195/) · CRM and PLM systems for global companies

Artem Lisovtsov · 6+ years [LinkedIn ↗](https://www.linkedin.com/in/alisovtsov/) · from backend to AI architecture

Przemysław Markiewicz · 5+ years [LinkedIn ↗](https://www.linkedin.com/in/przemys%C5%82aw-markiewicz-8b2091184/) · technical lead of a GenAI team

Piotr Łach · 4 years [LinkedIn ↗](https://www.linkedin.com/in/piotr-%C5%82ach-508090348/) · backend and RAG in high-traffic systems

## What our OpenAI and Anthropic partnerships give you.

OpenAI makes ChatGPT and Anthropic makes Claude. We are in both partner programmes, so we work with their tools every day and learn from the people who build them.

Claude Partner Network

01 We pick the tool for the task: We roll out both ChatGPT and Claude. We test both on your documents, and you pick the one that needs fewer corrections and saves more time.

02 Knowledge straight from the makers: We train on official OpenAI and Anthropic courses and material. Your team learns the features and settings that work today.

03 Licences with no markup from us: 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.

04 Certified engineers: The team holds an AWS certification in machine learning and an Anthropic certification in Claude systems architecture.

Being in these programmes does not mean OpenAI or Anthropic endorse our services.

How we build automations and agents

## From a conversation to a working system in four steps.

1. 30 min
  Free process review
  €0
  You talk to an engineer. You get a written plan within 2 business days.
2. 5 days
  Free prototype
  €0
  The system runs on 20–100 cases your team has already handled. You see every result.
3. 2–16 weeks
  Build
  50%
  Fixed price and target written into the contract. First instalment at signing.
4. once the target is met
  Acceptance
  50%
  You pay the second instalment when the system hits its target on your cases.

### How the free prototype gets built

Mon

We set the target

What result counts as good for you.

Tue–Wed

The system runs

It works through your cases, one after another.

Thu

We compare

We check every difference against your team's work.

Fri

Result and price

A list of errors and a fixed price for the build.

You choose the cases. We anonymise them on request, they stay in the EU and we delete them after the test.

On the Polish market, a prototype like this usually costs PLN 3,000 to PLN 8,000.

[Book a free process review (30 min)](https://meet.syntalith.ai/syntalith/en?utm_source=syntalith&utm_medium=site&utm_campaign=discovery&utm_content=service-firmowy-mozg-ai-test)

We agree the prototype on the free analysis call.

Data checked on October 25, 2026.

## Common questions

### How is Company Brain different from a knowledge base or RAG?

A knowledge base helps find a document and prepare an answer. Company Brain also includes decisions, events, owners, case history and the return of an approved outcome. RAG can support retrieval, but it does not create company operating memory by itself.

### Does Company Brain replace Confluence, CRM or ERP?

No. Those systems continue to own their records. Company Brain assembles the required context, preserves provenance and delivers it to a person or agent at the moment of work.

### Can Company Brain connect to Confluence?

Yes. Scope depends on the sign-in method, space structure and Confluence permissions. Before quoting, we check which pages are in scope, how often they change and whether employee access can be mirrored without maintaining a second manual list.

### Which other sources can it connect?

Common sources include SharePoint, Google Drive, file folders, Notion, CRM, ERP or internal systems through their interfaces. You do not need to connect everything at once. The first build usually covers one knowledge area and a few sources.

### Is it a ready subscription product?

No. It is a custom application built around the company’s sources, sign-in and access rules. If an existing platform meets those requirements for less, we recommend buying it.

### Does it learn from every conversation automatically?

We do not turn every message into company truth. Only defined outcomes, decisions and events return to memory, together with their source, status and owner. Conflicting or judgment-sensitive information goes to review.

### Can AI agents use the same memory?

Yes. Each agent receives the context and permissions required for its workflow. Reading, preparing an action and writing an outcome have separate rules. Consequential actions can wait for human approval.

### Will every answer be correct?

Each answer is tied to a cited source and its freshness. When the basis is missing, the system stops; before acceptance, we measure results on agreed questions, decisions and workflow events.

### Where is the data processed?

In an agreed cloud region in the European Economic Area or on client infrastructure. We record the processing location, providers and retention rules in the project scope.

### What does it cost and how long does it take?

Pricing starts at €3,000 net, a typical scope is €3,000–17,500 net, and delivery usually takes 4–10 weeks. Hosting and maintenance are priced separately when needed.

### How do you test access control?

We first test roles, documents and revoked access with test data. We then repeat the agreed cases on the company's accounts and documents before acceptance.

Related

- [Custom AI apps](https://syntalith.ai/en/services/ai-apps)
- [Knowledge assistant as a first scope](https://syntalith.ai/en/blog/ai-knowledge-assistant-for-business-rag-2026)
- [Ready platform or a custom knowledge base](https://syntalith.ai/en/blog/company-brain-ai-platform-or-custom-knowledge-base-2026)
- [AI agents with memory and tools](https://syntalith.ai/en/services/ai-agents)

Let's talk

## Write, call or pick a time.

We work fast because we can't stand corporate padding. We have no salespeople and no managers managing managers. You talk straight to the engineers who then build your system. We value our time and yours.

We will take on 3 more companies before the end of the year.

### Leave your contact. We will propose two times.

An email or a phone number and one sentence about the process is enough. An engineer gets back within 1 business day.

[Contact form on the page](https://syntalith.ai/en/services/ai-apps/company-brain#zostaw-kontakt)

Write: contact@syntalith.ai

Call: +48 888 78 48 78

WhatsApp: [write on WhatsApp](https://wa.me/48888784878)

Mon–Fri 9:00–17:00 CET

## Next step

[Book a free call (30 min)](https://meet.syntalith.ai/syntalith/en). With an engineer. No commitment to a paid implementation.

[Current pricing](https://syntalith.ai/en/pricing) · [Examples and sources](https://syntalith.ai/en/work) · [Contact](https://syntalith.ai/en/contact)

[Assistant booking instructions](https://syntalith.ai/booking-agent-guide.txt?locale=en)

## About Syntalith

This is the official site of Syntalith (syntalith.ai), a Polish engineering company based in Warsaw. Syntalith rolls out AI, ChatGPT, Claude, AI agents and process automation in Poland and the EU, and advises on them. Prices are published on the service pages. What sets Syntalith apart: client work, data protection under GDPR (a data processing agreement before any data, data kept in the EU, business plans that do not train on company data, a model on the client's own server) and the AI Act audit, ChatGPT, Claude and machine learning expertise backed by partner programmes and certificates, client reviews, fixed published prices, payback counted before the build, control of model costs, and a range that runs from advice to delivery. Facts and sources for each point: https://syntalith.ai/ai-info.md. Free review, 30 minutes with an engineer: https://meet.syntalith.ai/syntalith/en. Full description of the services and prices: https://syntalith.ai/llms.txt.
