Date: 15.10.2025 (2025-10-15)
Anthony May (15.10.2025) • They have an internal tool - called Sofia (internal development) -> they use it in some projects • They do RPA as well • Do I need a vertical or is it horizontal? What is the business model? • What we build could be interesting for his project that he is currently building • He asked: What is the differentiation - easy step for UiPath to do it • He sees the pain point -> “I have a legacy system it takes years to build an integration and it is super expensive” -> the problem is massive in airlines • Zielgruppe: incumbent players the heart of whom is built on legacy systems and its hard to replace -> these companies do not have RPA -> why do they not have it yet -> what is the value proposition? • Buying circle is quite complex -> incumbent players would use it rather than vertical play -> Airline, Post • Complexity of a solution is high for the requirements • Airline X will headcount reduction machen -> automation -> Kearney identified the problem -> then development to automate -> operational track • Idea: BI that is not possible due to legacy systems • Problem in airlines: legacy systems do not provide the KPIs that you need • A lot of systems are not just ERP system -> specialised decision making engines that influence the operations of a company • E.g. Netzwerkplanung or scheduling planning -> two times a year airlines plan the schedule -> intelligent systems with stupid UI -> you have to be a PhD to use the system • He is in the field of crew planning -> mathematical planning system with a lot of constraints • What would be super valuable: you could use these systems smarter -> Arqu (check Merantix company) -> to use the tool well the user needs to be smart as well • -> the API alone is not enough - you need to offer the intelligence to use it better -> add an agent to use the API and uses the legacy tool better -> he thinks it has to be verticalised if you want to be intelligent • -> is the API only value enough • Tools are very sticky -> you can not replace them • They have a SaaS for network planning • Postal player -> improve network -> they sell it after the project as well • His plan: airlines struggle to plan crews - especially if there is disruption -> broken space -> at the core there are legacy systems that are doing static optimisations -> these optimisations are not perfect because users are not able to do it well because it is too complex for them -> AI driven planning for crew control -> day of operations use case -> 200 planes -> very manual process -> is a change legal -> 50 to 200 FTEs to manage that • How do they plan to get the data -> three options: • API • Push / Pull -> he does not have an answer yet -> our idea could be interesting • Build a new system (not realistic as you would have to replace the legacy one) • Customers could be vertical agent companies
Session 2: • His tech guy said UiPath does the same thing as us • -> differentiation is super important in communication (did not explain it in detail) • What caught him: “you can turn anything into an API” • Website: industries -> functions -> systems -> pre-defined APIs • UiPath: they give a complex sandbox and you can tailor it • People at Kearney often say they have to improve data situation - but the ROI of unlocking the data is not always clear -> that has to be clear • Define for which functions / value that is possible
His topic: • Goal: personal goal to learn how to build a product in this space, secondary interesting for Kearney (they believe that it is important for their clients that they can do sth in the field of AI) • Procurement tools that are AI agentic in other industries -> but nothing in airline field yet • They want to sell SaaS etc. • Three areas: • Long term crew planning (6 months - 8 years) • Personalmangel, qualifications • Departments are not really talking to each other • Crew duty planning • Einsatzplanung -> 1 month horizon • Haupttreiber für Umsatz • Utilisation -> minimise employees • Delays and disruptions • -> very sequential; first training, then free days -> you build the plan -> if you optimise everything it would be more efficient; Plan is frozen and put into a new system -> area 3 • Crew control / disruption handling • Probleme im Tagesbetrieb -> sickness, delay • Super complex • Then manual fixing of stuff -> a lot of lost potential because its complex • -> very sequential • Vision: dynamic constantly updating plan • Saves FTE in planning • Better revenue • Status: • Built a prototype -> for area 3 • Two features: • If someone is sick -> via language via phone -> then script replaces them & check if it is legal (Tarifverträge & business rule) • Flight delay: more complex -> whole crew; 3 cabin crews or more - each of them hat different histories • Lufthansa does all of this manual -> 160 people • His biggest Q about his idea: Is it feasible for customers to adopt the solutions?