CortexMine has been accepted into KPT ScaleUp Booster. Six months of intensive work with Polish manufacturers, with mentors who know factories from the inside, and with a team that has been running this accelerator for seven years. We are writing about it because it is a concrete element of our trust trajectory and a concrete moment on our path to scale.

KPT ScaleUp Booster is run by Krakow Technology Park, Poland's largest Industry 4.0 accelerator. Seven completed six-month editions, 64 technology companies supported, more than 25 industrial partners, up to PLN 300k of funding per company and two verticals, industrial and gaming. For us the number that matters is not the grant. It is the access: large manufacturers who are actually willing to test a solution on their own documents.

What KPT ScaleUp Booster is

KPT ScaleUp Booster is an acceleration program run by Krakow Technology Park, funded by European Funds for a Modern Economy (PARP, FENG). It is the largest Industry 4.0 accelerator in Poland, with over 100 alumni and a clear focus on digitalization, process automation, and AI in production and maintenance.

Among its industrial partners are companies such as Grupa Azoty, ASTOR, Werner Kenkel, Maspex, and MAN. This is not a logo wall. These are organizations actively looking for concrete technology solutions to deploy inside their plants. A six-month structure, a grant of up to PLN 300,000 for product development, and mentoring from people who know factories from the inside. Frankly, the grant is not the main reason we are here.

Why we fit this program

We are building private AI for European manufacturers. On-prem. Source-grounded. Audit-ready. Three productized workflows: a service and troubleshooting assistant, work instruction generation, and technical offer preparation (including drawing-to-offer). The entire working infrastructure stays inside the factory perimeter. No public cloud. In the NIS2 reality and the amended Polish cybersecurity law that came into force in April 2026.

Mid-sized Polish manufacturers (50 to 500 FTE: metals, mechanical, electrical, automotive subcontracting) need concrete AI workflows that pay back within 12 to 18 months and pass a CISO audit. They do not need "AI for everything". They need tools for the problems they actually have: quoting cycles that take too long, scattered technical documentation, service desks that cannot keep up. Those are exactly the problems that KPT ScaleUp Booster targets through pilot deployments.

What we are doing here

  • Direct pilots with large Polish manufacturers. The program targets validation in real production conditions, not on slides. This is exactly the stage we are in.
  • Access to technical decision-makers in Polish industry. IT managers, operations directors, engineering office leads, security officers. People who have seen several AI hype cycles and want to know what actually works, what it costs, and what happens to their data.
  • Mentoring from a team that knows factories from the inside. Seven years of accelerating in this category is not a marketing line. It is concrete know-how on what not to do in the first 90 days of working with a manufacturer.
  • Polish regulatory and financial framing. NIS2, the amended UKSC, AI Act, GDPR. The program knows these realities and translates them into deployment decisions.

First weeks: conversations, people, first hard numbers

What has already happened, short and concrete:

  • Sessions with the KPT team and with the other startups in this cohort. A few conversations that are already changing how we plan the next quarter. We met people we would normally meet only six months from now, after three networking rounds and five cold emails.
  • First mentor sessions on the product, sales, and regulatory side. Discussions about how a typical pilot differs from Co-Development with a KPI guarantee, about the buying cycle in Polish manufacturing, about how to structure the first 90 days of work with a factory so that it does not end as "an interesting POC with no follow-up".
  • Direct conversations with industrial partners of the program. This is the moment where the dialogue starts with a specific production problem, not with a generic pitch. The questions are technical, the requirements are hard, and data sovereignty and NIS2 are on the table from the first minute.

We are not naming the companies we are speaking with. These are early conversations and we respect the partners' right to privacy at this stage. But we can confirm: the context of these meetings is exactly what we expected. A specific problem, specific data, specific compliance requirements.

We are glad to work with industry directly

This matters to us. Building private AI for European manufacturers is not a project you can run from an office while reading reports about "the AI revolution in factories". You need to be in the factory. You need to talk to people who spent ten years on the production floor and know precisely why "AI for everything" does not scale. You need to see technical drawings from 1997 that nobody wants to digitize. You need to understand why a Polish CISO of an organization in scope of NIS2 will not let ChatGPT Enterprise inside their perimeter, even if they wanted to.

KPT ScaleUp Booster gives us all of that in condensed form. Direct contact with manufacturers. Concrete pilots. Mentoring from people who do not sell illusions. This is exactly the context in which we want to build for the next six months.

Why we are writing about this

Two reasons. First: it is the practice we also recommend to our customers. Be transparent about your partners, your regulatory framework, the accelerators you participate in. A CISO of a key or important entity under NIS2 needs to know who they sign a contract with. KPT ScaleUp Booster is a concrete, verifiable element of our trust trajectory, alongside Microsoft for Startups and NVIDIA Inception.

Second: if your manufacturing company is considering AI outside the public cloud, this is a good moment to talk. We have more bandwidth for concrete technical discussions, more access to the Polish industrial ecosystem, and a stronger pilot structure than we had three months ago. If you are interested in drawing-to-offer, an AI service assistant for manufacturing, or generating work instructions from your own documentation, let us know.

What comes next

Over the next five months we will publish notes from the program (without leaking confidential details, but with observations that might be useful to others). Topics we know will come up:

  • How a mid-sized Polish manufacturer actually evaluates an on-prem AI deployment in 2026.
  • What "audit-ready" means in practice, when the CISO is on the other side of the table.
  • Drawing-to-offer in a real Polish engineering office: where it works, where it bites.
  • How to structure an on-prem AI pilot so that it does not end as "an interesting POC with no follow-up".
  • Digital Workers versus the seat-based pricing model in a mid-sized manufacturer: when each option actually pays back.

If you want to discuss your workflows, your NIS2 audit, or simply see CortexMine live: book a 30-minute call with the founder (no pitch, just the problem).

And a thank-you to the Krakow Technology Park team for the invitation to the program. Seven years of experience is clearly visible in how this accelerator is run. We are glad to be part of this cohort.

Fryderyk Pryjma, founder, CortexMine

Prefer to check on your own first? The readiness mini-audit takes 10 minutes and leaves no data behind.