TRAXIA case study

Case study – rolling out AI at TRAXIA

Time: 60 minutes Teams: 3 teams of 3–4 Output: a completed AI rollout canvas + a 4-minute pitch to the board

You are a management working group at TRAXIA s.r.o. (intelligent transport systems, ~220 employees, HQ in Banská Bystrica, branches in Bratislava, Žilina, Košice and Kraków). CEO Martina Kováčová wants a proposal by the end of the month on how to introduce AI in one process. The company runs Microsoft 365, and managers have Microsoft 365 Copilot licences. All names, companies and figures are fictional.

You are a sceptical CFO and a risk auditor. Read our AI rollout proposal [paste the canvas text]. Find the 5 weakest points: unrealistic business-case assumptions, missing risks (GDPR, AI Act, security), weak spots in the pilot and adoption. For each, suggest how to fix it.

Case A – Service 24/7, East region (Košice)

Process owner: Marek Oravec, head of Service 24/7

Process today: A customer (road authority, city, railway) reports a fault on a device – a variable message sign (VMS), a WIM scale, level-crossing protection or a junction controller. The dispatcher creates a ticket in TRAXIA Servis, sets the priority and assigns a technician. The technician goes on site, diagnoses, repairs or orders a part.

Pain points: Manuals are PDFs in three places (SharePoint, file server, technicians’ phones), some only in German and English (supplier SENSOTEC). New technicians phone experienced colleagues. Fault history has been in the system since 2021, but nobody searches it. The customer RSC Stred has filed a complaint.

Constraints: Technicians work on the carriageway and near tracks – safety procedures must be exact. Hiring new technicians is frozen for 2027. The service system has an API, but IT has only 20 days of capacity a year.

Case B – Tenders and bids

Process owner: Lucia Tóthová, sales director

Process today: Sales monitors public-procurement calls, decides bid/no-bid, prepares the technical part, the qualification documents, references and the price. The managing director signs the bid.

Pain points: Tender documents often run to 200+ pages. The library of past bids (~300 documents) sits unstructured on a file server, references in Excel, certificates in e-mails. The company doesn’t yet have ISO/IEC 27001 (audit 03/2027) – some tenders require it. Sales would like to submit more bids but lacks capacity.

Constraints: Prices and calculations are confidential. Deadlines are fixed. An error in a bid = disqualification.

Case C – Technician onboarding

Process owner: HR (2 people) together with Service 24/7

Process today: A new technician receives a 120-page handbook, completes mandatory training (safety when working on the carriageway, electrical qualification) and shadows an experienced colleague (“buddy”) for 3 weeks.

Pain points: Newcomers keep asking the same things. Experienced technicians lose time and get frustrated. Polish colleagues only have Slovak materials. HR doesn’t know where newcomers struggle most.

Constraints: Employee personal data. AI used to evaluate people’s performance = high risk under the AI Act – don’t use it to assess newcomers. Safety training must remain certified and led by a person.

Case D – Project reporting for the board

Process owner: Peter Hudec, CFO

Process today: Every month the controller assembles the board report from ERP (project costs), Excel plans and project managers’ e-mails. The report contains tables and written variance comments.

Pain points: 80 % of the comments are the same every month. Warning signs (late supplier SENSOTEC) sit in e-mails, not in numbers. The report is ready only on the 8th working day.

Constraints: Numbers must be exact (review-ladder level 2). Few people have ERP access. Financial data is confidential.

AI rollout canvas

Team: ………………………. Case: A / B / C / D

Case-study scoring

Each criterion 0–2 points, max 12 points, pass ≥ 8.