Stefan Schlossarek

Date: 11.11.2025 (2025-11-11)


Stefan Schlossarek (11.11.2025) • https://www.linkedin.com/in/stefan-schlossarek-59b936a/ • Self employed as CFO / Finance Manager • Per Du • Via Teams • My message: • Wir entwickeln proprietäre KI-Modelle für Controlling & Accounting, die auch für komplexe Fälle trainiert sind - etwa mehrere Purchase Orders auf einer Rechnung, Price-Variance-Tolerances bei Lieferanten etc. • Ziel ist eine ~80% Automatisierung in AP/AR und auch 50% weniger Aufwand beim Closing & Reporting - ohne neues SaaS-Tool, direkt integriert in bestehende Systeme (auch ohne API) und angepasst an aktuelle Prozesse.

Notes: • Context: • Companies: CFO for Cybersecurity company -> they were in an acquisition by a Swiss company (friendly takeover) -> CEO was sick -> they had someone that does the Jahresabschluss etc. und Leitung der Übernahme • Structure controlling, Liquiditätssteuerung • Before Playmobil -> Controlling, 10 people, new Planungstool • Worked at Horst Brandstätter Group -> intro? • Automation: • • Biggest objections / challenges in his daily life? • Kostenstellen und Kostenträgerrechnung -> invoice comes in -> which project, which Kostenstelle, does it have to be split? • Important to do Profitabilitätsrechnung per product etc. • A lot of mistakes are done -> data is wrong -> garbage in / garbage out • Huge field that is needed in every field • Partly in Buchhaltung, partly in Rechnungsprüfung • Vorkontierung partly automated -> but does not really learn yet • Very rule based • Budgetierungsprozess • Mainly done on data of the past • Planung is done auf Kostenarten • Would be nice to have it based on Kostenstellen • E.g. new employees -> higher phone costs • Dear AI: here are the costs across last 3 years -> what happens if I hire 1 employee in Produktion • Office space, PC, phone, car, expenses per function, Jobrad etc. -> take average • There are planning tools in the market • Bürogebäude sind Sprungfixe Kosten • I think we need to build Excel files in DB style and let AI iterate the model with the human • Human should review in Excel • There are Planungstools that are doing it quite well: • Most companies do not have planning tool but most have ERP • Often ERP systems have internal planning tool • Check ERP APIs • Often companies have a tool but do not really use it • When he leaves the tool dies again -> he finds an internal champion • Examples: • Quantum (Thinking Networks) is open Source afaiu • EcomPlanner.de (for SMEs and startups) • He talk to someone that works well with ACP Digitalisierung -> they do a lot of AI • They tried upload data into an LLM and it was really good -> good suggestions -> he can do intro • Erik Purwiens or sth. - I can contact him and mention Stefan Schlossarek • He knows them well • They also buy companies -> often in a difficult situation -> 20 Systemhäuser (MSFT, Cybersecurity) • Radiology contact (reinvented CT): Dr. Oliver Heid • https://www.linkedin.com/in/oliver-heid-893927154/ • Profitability management • Many companies do not know where profit comes from • They do not know where they make losses -> dumping prices to acquire them -> later more revenue -> most of the time it does not work • Introduced profitability management at many customers • They need profitability on product level, company level, Sales manager level etc. • Current blocker: Zuordnung der Kosten to the right customer or product • Direct costs are relatively easy • E.g. service technicians are tricky • Deckungsbeitrag 1 etc. -> understand these finance metrics • How he did it: via Stundenrechnung (ETSM Systems) -> at the beginning it is a process change • Maybe we can proxy it with some logic? • Side effect: you can realise that there are too many employees -> you can save costs as well • When technicians are sloppy: • Make bonus dependent on data quality • Talk to employee to explain it to them • Phone times can be done automatically via the phone system • Pre-Sales: • They also document their time • You can estimate it • In reality actuals are double vs. budget -> Nachkalkulation is relevant and compare with Vorkalkulation -> comparison • He implemented it with Excel • Only one time he did with ThinkingNetworks • Vollkostenrechnung etc. • Even negative costs cover some fixed costs -> killing the worst customer but then the next one is the worst • He is willing to give feedback in the future • Most valuable flows for him? • • Context: team setup? internal and external? • • Which tasks are the biggest pain? With magic wand -> what would he solve? • • 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? •