PohlCon: Hours of catalog work turned into minutes per quote

PohlCon is the Czech branch of a German manufacturer of construction systems, and its sales team compared and matched products from different suppliers to its own range by hand. Mapping supplier products onto the matching internal items, including technical parameters and current prices, took up to three hours per catalog. Preparing a quote took 20 minutes for a small order and up to 180 minutes for a large one, always with the risk of a typo or a wrongly chosen alternative.

The deeper problem was who could do the work. The whole process rested on the senior experts in the team. Their know-how was documented nowhere, and juniors could not get to customer contact without technical preparation. PohlCon needed a process that works independently of specific people.

PohlCon originally approached us with what was essentially an "implement AI for us" request. We started with an analysis of how the team actually works instead - and it showed that a large part of the potential savings needed no AI at all, just well-designed automation. So that is what we built: an automation layer connected to the company CRM. Internally, the team calls it "překlápěcí tabulky" - their "flip tables", because it flips supplier catalog data into a PohlCon quote.

AI has exactly one job in the system, the one it is genuinely good at: mining structured data out of supplier catalogs, with their thousands of items and technical parameters, into values the tool can work with. The quoting itself is deliberately plain, reliable code that does the same thing the same way every time. A salesperson picks what the customer needs, product variants are ready in milliseconds, and a complete quote with current prices from the connected CRM takes minutes.

It would have been easy to push AI into every step of quoting and call the whole thing an AI tool. We did not, because it would be slower, more expensive and less reliable. Not every problem is an AI problem - the goal is efficiency, not AI for its own sake.

The output is a quote without typos and without wrongly chosen alternatives, and a junior salesperson can produce it without detailed technical knowledge of the product range. Senior time moves from administration to customers. And nothing here is specific to PohlCon's systems - the same automation connects to whatever CRM or ERP a company already runs, including systems managed by an external vendor.

The first results came within a few weeks in the form of a quick fix; the full solution was then built over the course of 2025. After rollout, quote preparation time dropped by 75 percent on small orders (from 20 minutes to 5) and by 94 percent on large ones (from 180 minutes to 10). The five-person sales team frees up roughly 220 hours a month. At an internal rate of 400 CZK per hour, that is a potential saving of about 88,000 CZK a month, against running costs of around 1,000 CZK - the investment in development paid for itself within a few months. The math is not subtle.

Team productivity at least doubled, often tripled. Instead of administration, the team now spends its time building customer relationships, and what customers notice most is the speed and accuracy of responses to their inquiries.