Routes and documents in one system

We publish our prices up front, net of VAT: automating a single process is from 8,000 PLN, and a CRM or ERP system with AI is 20–90k PLN. In the Pricing section you can put together a rough estimate from modules. The exact price is set individually after a free consultation, because your processes decide the scope.

No, and we will not pretend otherwise. We build systems with AI, integrations and document reading for other industries, and the routing engines we propose are open: their code and documentation are public. That is why we start with a free consultation and a small stage that ends with a sign-off: before the whole thing exists, you see a working piece on your own data.

They are open programs that we can deploy on your server. VROOM works out the delivery order for many vehicles: with time windows, capacity, driver breaks and pickups along the way. Valhalla and OSRM calculate routes and travel times on OpenStreetMap roads, and Valhalla also takes a truck’s height, width and weight into account. Google OR-Tools and Timefold Solver Community edition are solvers for rules that a ready-made optimizer cannot handle. VROOM and OSRM are under the BSD licence, Valhalla under MIT, OR-Tools and Timefold Community under Apache 2.0, so you pay no licence fee for them. Paid are, for example, GraphHopper route optimization in the cloud and Timefold Enterprise: if they turn out to be needed, we will tell you before the quote.

No. The routing engines work on OpenStreetMap data, released under the open ODbL licence. There is one condition: the map carries the credit “© OpenStreetMap contributors” with a link to the OSM copyright page. An open geocoder, such as Nominatim or Photon, turns addresses into coordinates. You pay for the server on which the engines and maps run. How precisely the map describes truck restrictions on your routes, we check on your own data before we promise a result.

No. The system adds what yours do not do: route planning, document reading, a client portal. If your TMS or WMS has an API or imports from a file, data can flow without retyping. What they expose, we check at the free consultation.

It can get it wrong, especially with handwritten entries and a poor photo, which is why we do not leave it the last word. AI shows the fields it read next to a preview of the document: a person checks them, corrects them and approves them before they reach the order. We use ready-made models and do not train our own.

If our work gives us access to personal data, for example of drivers or recipients, we sign a data processing agreement before we start (Art. 28 GDPR (RODO)). Access inside the system is limited by roles: a client sees only their own orders, a driver only their own runs. The connection is encrypted (SSL), backups run automatically, and the test environment is kept apart from production. Which AI services we use and what data reaches them is agreed with you before launch.

The timeline depends on the scope and we give it together with the quote, broken down by stage. Every stage ends with an acceptance report, so you see a working part of the system, for example route planning on your own orders, before the whole thing exists.

You will. You pay for the work, not for a licence: there is no subscription. When we finish, we hand over the source code, the documentation and the access credentials, so you can develop the system yourself or with another company. The open engines stay under their own licences, and we describe their terms in the documentation.