Kundberättelser
Medidesk
Healthcare SaaS · Poland · 2026

More time selling. Same team.

Medidesk driver patient-klinik-kommunikation för 1000+ vårdenheter, använder redan AI inuti produkten. De ville ha AI inuti företaget också. Vi tog in extern data och signaler. Säljteamet lägger nu tid på ICP-anpassade kunder redo att prata, inte manuell prospektering.

Medidesk customer story
Spela upp video
3:37
2.5×

more time selling to ICP

AI absorbed the repeatable pre-sales work

1 week

from kickoff to a working first version

prototypes in hand, not decks

30 → 1

vendors pitched, one got the work

picked on behaviour, not promises

Daniel Nowocin ran an open call for an AI partner. Thirty vendors pitched, three made the shortlist. What decided it was not the strongest deck but how the conversations went: responsive, concrete, small steps forward. Within days of starting, he had a clickable prototype in his hands.

If you are choosing an AI partner right now, his filter is worth copying. A deck can promise anything, so Daniel judged behaviour instead: how fast the answers came, how a vendor handled a change of direction, whether the conversation showed people who build or people who sell. Experience and references counted, but behaviour decided.

Medidesk already runs an AI-first product, patient-clinic communication for more than a thousand medical facilities. The decision at the start of the year went further: bring AI into the running of the company itself, with every step proven as a small POC before committing to a full build.

The first system landed in sales. Before a first call, the brief is already there, built from outside data and signals: how big the clinic network is, who runs it, what they use for patient contact today, and which signals say they are ready to buy. Two minutes of reading, then the call. Selling time goes to those conversations, not to manual prospecting.

None of it started with a model. All of it started with ordered data.

That is Daniel’s own rule, and ours. Before any build, we defined the scope of input and output data together and got it structured, so the AI reasons over the company’s reality instead of guessing at it. It is the unglamorous step most AI projects skip, and the reason these systems fit how the business actually works.

From there the pattern repeats: a small stage, shipped in days, verified on the market, then extended. Each working system opened the way for the next.

“For me it matters that the partner, beyond the technical knowledge, feels the business, feels the logic of building a business. That join of technology and business understanding is the key for me.”
Daniel Nowocin
Founder & CEO, Medidesk

What moved to the system. What stayed with people.

Daniel drew the line himself: technology takes what needs automation, people keep what needs a person, and the team puts its time where it earns the most.

The system
What AI now handles
  • A ready brief before every first call
  • Outside data and signals, in one picture
  • Qualifying who fits the ICP and who is ready to talk
The people
What stays human
  • Client conversations, from first call to close
  • Judgment calls and business decisions
  • Quality assurance on what the systems produce
  • The contact that has to stay human

In Daniel’s words: what used to be done by a person is now done by AI that supports the whole process. People focus on quality and on the contact that has to be human. The technology takes the rest.

The playbook to steal

Daniel’s advice to another founder is the playbook, and it has less to do with buying AI than with how you build.

  • Split the elephant. Don’t build huge solutions. Divide the project into smaller stages, ship one, verify it on the market, then extend it in the next iteration.
  • Put the data before the model. Define the scope of input and output data first, and get it structured. Every project that ends up correct starts there.
  • Expect something you can test. If a partner can’t put a working prototype in your hands before the big commitments, you are buying a promise, not a capability.
  • Pick a partner who feels the business. Technical knowledge on its own is easy to find. What makes the work fit real needs is technology combined with the logic of building a business.

Låt nästa framgång bli er.

Varje berättelse på den här sidan började med ett 30-minuterssamtal. Berätta om er situation, så berättar vi hur det första steget skulle se ut.

Boka introsamtal