We state prices up front.
Process scan €0 ($0), implementation specification from €1,200 (approx. $1,300), automations from €3,500 (approx. $3,800), apps and agents from €6,000 (approx. $6,500), training from €1,200/day (approx. $1,300/day), maintenance priced individually. All prices net. If you commission the system build from us, we credit the specification fee toward the build.
- A 30-minute call with the engineer who would lead the work.
- We review the process, its exceptions, and its current cost.
- Within 2 business days, we send a written summary.
Selected client work
From a personal AI agent and training to systems used by entire teams.
Packages
What you buy, what it costs, and when it's ready.
Builds typically land between €6,000 and €35,000 (approx. $6,500 to $38,000). Each level has a clear entry price. You get a fixed quote for your process in a proposal after the free scan, before you sign anything. The written scan takeaway is built to forward to whoever signs the budget.
You start here
Entry · start here
Process scan
30 minutes with an engineer about the processes that cost you the most. Within 2 business days you get a written summary: what to automate, in what order, and what it costs.
30 min + summary within 2 business days
€0($0)
Book a slotAfter the scan, you pick the path
Path A · a free proposal with a fixed quote, within 5 business days, valid for 30 days.
Path B · an implementation specification, when you want a clear decision and a fixed quote before building.
Path B · optional
Implementation specification
A process map, architecture, plan, and fixed quote, followed by a documented recommendation on whether to build. If you commission the system build from us, we credit the specification fee toward the build.
5 business days
from€1,200(approx. $1,300)
See scopeBuilds and training
Work off your team's plate
AI Automations
Processes that run themselves: inbox, documents, CRM, reports, and system-to-system integrations.
2–6 weeks
from€3,500(approx. $3,800)
for simple, rule-driven processes
See scopeWork off your team's plate
AI Agents
Systems that run a process themselves: routine steps close without a human, edge-case decisions escalate to you, every move is logged for audit.
6–16 weeks
from€6,000(approx. $6,500)
See scopeA tool for your team
AI Apps
A tool used by your team: document search, data extraction, guided work, and an internal panel.
4–10 weeks
from€6,000(approx. $6,500)
typically €6,000–14,000 (approx. $6,500–15,000)
See scopeFor owners and executives
Personal AI agent
A personal operator for owners and executives: briefings, research and decision prep within set boundaries, on infrastructure you control.
narrow scope · usually a few weeks
from€1,200(approx. $1,300)
operator on your infrastructure, data in the EU
See scopeTraining
AI Training
Closed corporate training in safe, effective AI work, taught by the engineers who deploy these systems.
1–2 days
from€1,200/day(approx. $1,300/day)
See scopeCourse · 1:1 or company
AI-Native course
A one-to-one or company course on your codebase. You keep the working method, templates, and tooling created during the course.
1:1 or company · on your repository
quoted after a call
paid course · net
See the AI-Native course scope
AI apps are operated by your team; automations and agents do the work without it. We decide the form after the scan. All prices net.
What you receive at these prices
Cost-invoice coding
In the recorded run, model cost was USD 0.004514 per invoice. When an invoice lacks supporting evidence, it goes to the accountant for clarification instead of becoming a ready posting.
Quote-request pricing
In the recorded run, all 40 quote lines had a named price source. A line without one is held for clarification.
Case-law research
The models run locally; the legal team's questions incur no external API fees.
What moves the price
Three things set the quote.
- Data sources and integrations
- How many systems have to agree with each other: inbox, CRM, ERP, documents. Every integration has to be built, tested, and maintained.
- Autonomy and risk
- The more the system decides on its own and the more its mistake costs, the more work goes into boundaries, escalation to a human, and tests on your data.
- Audit and compliance
- A trace of every decision, human sign-off on production effects, and mapping to AI Act requirements are in scope from day one.
Price follows the work in scope: data sources, integrations, decision autonomy, risk, and audit requirements. The scan determines whether that work needs an automation, an app, or an agent.
Result guarantee
One measurable target, written down before the build.
Before the build, we agree one measurable target and put it in the contract. A pilot on your data usually takes 3–6 weeks. If it misses the target, we complete the contractually agreed fixes within the pilot's fixed price before the full build proceeds.
A gate that failed first
This guarantee is backed by a measurement history that first missed its target and was published anyway. Resident-ticket routing reached 0.828 against a 0.850 threshold; after remediation, a new run reached 1.000. The case preserves both results and the operator-approval boundary.
Payment and scope
Questions before you buy.
How does payment for a build work?
Builds are usually split in half: 50% on signing the scope, 50% on delivery. The implementation specification is paid once, up front; maintenance is billed monthly. We start work after a signed scope and the first payment has cleared.What does the starting price include, and can it change after signing?
“From” is the floor for the simplest, rule-driven process, with no catch attached. You get a fixed quote for your process in the proposal after the scan, valid for 30 days, and you sign scope and price together before we start. The price doesn't move mid-build unless you change the scope, and then we show the new quote for sign-off first.Is maintenance cost, including AI models, predictable?
Yes. Model usage is charged in addition to our fixed monthly fee. We meter it per case, report it, and can set a monthly cap. We estimate your cost from expected volume during the scan. In the recorded test run for data questions, model cost was USD 0.004400 per question.What if we want to leave or switch vendors?
At handover the code, prompts, eval data and documentation pass to you: the system is yours, you don't rent it from us, so you can run it yourself or with any team. The implementation specification is a portable document for any vendor, so nothing locks you to us even at the start.Do the prices include VAT?
No. All prices on this page are net: the invoice adds 23% VAT (reverse charge applies for EU companies outside Poland with a VAT-EU number).Does the process scan commit us to anything?
A 30-minute conversation and a written summary within 2 business days. It ends with a recommendation; when automation is not justified, we put that in the document.Does the specification fee count toward the build?
Yes, if you commission the system build from us. We then credit the specification fee toward the build. The document remains portable, and you can still hand it to another vendor.How much does the implementation specification cost?
The implementation specification starts at €1,200 net (approx. $1,300). If you commission the system build from us, we credit its fee toward the build. The other entry prices are listed on this page.What about scope beyond €35,000 (approx. $38,000)?
Enterprise builds (several connected systems, regulated environments, budgets of €47,000–120,000 (approx. $51,000–130,000), RFP processes) are priced individually. A call confirms whether the work fits our team.
30 minutes, a written summary, zero commitment.
You can inspect 28 system descriptions before booking: client work and tests on test data. Each entry states how it works, the build cost, where a person decides and what has been measured.
- A 30-minute call with the engineer who would lead the work.
- A review of the processes that cost you the most time and money.
- A written summary of what to automate first and the likely cost range.
Times are shown in your own time zone. We work with clients across time zones.
The scan chooses one process to assess, and within 2 business days you receive a recommendation, including when a simpler route is the better fit.