Heinz Meier

Date: 24.11.2025 (2025-11-24)

Heinz Meier (24.11.2025) • https://www.linkedin.com/in/heinzmeier/ • CFO at Die Careum Stiftung versteht sich als Organisation, die den Dialog zwischen den Akteuren im Gesundheitswesen fördert (120 ppl) • Also Dozent and in real estate on the side • Zurich • Via Teams • Per Du

Notes: • Notes: • Where they use it: relatively little • They have some automations • AP inflow: digitalisation of invoice • Kontierung: • Zahlungsprozess: • ERP: Abacus -> they use AI by them • Erkennen von Lieferant, Zahlungsverbindung etc. 99% accuracy • Erstellen von Suppliers • Kontierung schwach • Standardkontierung hinter Supplier • Kostenstellen auch nicht wkl • They are not really beta testers -> they wait until Abacus features work • Bigger topic: workflow itself -> Abacus approach is the wrong one -> always a human to start the process -> external employee checks inbox and Veratbeitung 1h per day to assign invoices to different workflows (also sanity checks correction) • E.g. Kontierung is missing • Pure AI is used barely (e.g. ChatGPT) • CoPilot: Sitzungsprotokolle within Teams • Then person finalised it • ChatGPT: Textsuche -> Brief oder Email Antwort schreiben • E.g. KnowHow suchen • E.g. Mehrwertsteuergesetz -> faster than original search • Team: • 80% 1 FTE and he a bit • 250 people in total • They have a Verlag • They developed a Lern-Coach für Lernende basierend auf Lerndaten -> Avatar you can talk to (and write) - even recognises sentiment (happy vs. stressed) -> adapts the learning process to that -> you can ask what can be found somewhere -> multiple choice Aufgaben incl. solutions etc. • Works well but limited to learning • They do not have enough people to do it internally -> he wants to buy it from external • He wants to have a GPT to upload your PDFs and then it works • • Automation: • Real estate: • Störungsmeldungen die automatisch weitergeleitet werden • Zugangskontrollen: welche Türen wie lange offen etc. • Verband: Swiss Accounting -> tackle them • Revisionsgesellschaft they also hear that they think about AI • They all want to build internal GPTs • • Biggest objections / challenges in his daily life? • • Most valuable flows for him? • • Context: team setup? internal and external? • • Which tasks are the biggest pain? With magic wand -> what would he solve? • Firmeninterner ChatGPT • Add all documents Unterschriftsreglement, Spesen, etc. • Beispiel: Welche Feiertage, was ist Reglement • Always most updated regulations • Datenumwandlung von PDF in Excel • ChatGPT is already good • Automatic prep of input fields for Abacus • Reporting & analysis area • These topics take a lot of time -> much potential • Search for Belge -> root cause analysis • Anomaly detection -> suddenly high invoice • Forecasting • Not that much potential for them • Kosten sind auch relevant (800 CHF per month for ChatGPT is much) • Use does not represent the 800 CHF • A lot of tools pop up and then diminish again • They want to wait and then take one that is mature • Main value in reporting • Budgeting -> a lot of data comes from a system and then a lot of formulas in Excel -> a lot of manual work and error prone (creation of visuals) - then creation of reports • Excel: the size of data that can be managed is small -> quickly very complex • MSFT PowerBI tool is already there • Challenge get data in one place and connect it • Review Excel: is there a cell in which there is a number instead of a formula • Plausibilitätscheck and improve data quality; simple statistics
• Sometimes he needs to take a statistics book to understand • Copilot is more annoying than it helps • Tools they use: • Abacus ERP • ERP for departments -> Verlag mit Verlagslösung (Fiatos) • Shop where students can order Lehrmittel • CRM relevant data in there • Schulverwaltungssystem / ERP -> verschiedene Marken • Orders of students etc. • Important for analysis and forecasting • Number of students etc. • They have a CRM: MSFT Dynamics -> goal to have all data there but does not work well yet: different format and data privacy • Umfragetools: employee surveys (iCommit common in CH) • Evlautationstool of teachers and infrastructure • HR: Fluktuationsraten, Bestand etc. • How do they get data: • Either Excel templates or standard exports that he transforms into data format -> he does it manually or ChatGPT (saves a bit of time but not really good - some mistakes - he has to check) • 3 people in finance and HR; assistant from CEO • Pitch him cursor for FP&A - ask for example data • Which tasks would he focus on? • • Do they already have automations? Which ones? How well do they work? • • Which companies would he target? • • General feedback to pitch? •