
Travel information systems: 25 percent savings in cloud costs and stable operation thanks to FinOps
The project in brief
A distributed travel information system with many data sources and target groups was consistently aligned with the AWS Cloud. Our focus: cost transparency at team level, lean operating processes and resilient observability. The result: 25 percent lower ongoing cloud costs - and an operation that is prepared for growth, disruptions and changes.
Results in figures
Reduction of monthly cloud operating costs by 25 percent
less by reducing the availability zones in the development environment
less due to time-controlled shutdown of the development environment
Identified additional costs due to scheduler bug; lesson: establish early anomaly detection & monitoring
The challenge: autonomy, complexity and a lack of transparency
The original challenges of the customer project were primarily in the areas of autonomy, complexity and transparency. In the 25 teams, a high degree of freedom meant that each group established its own architectures, tools and operating processes, which made it much more difficult to manage the project uniformly. In addition, the incorrect sections of the AWS accounts prevented a precise allocation of costs and the Cost Explorer used was not able to provide a clear view of the individual teams - so there was no transparency regarding the distribution of resources.
Another obstacle arose from the chosen pipes-and-filter architecture, which built in numerous intermediate stations in the data flow and thus drove up network costs disproportionately. In addition, three separate environments for development, testing and production were each operated via three availability zones (AZs); the development environment in particular ran around the clock, causing ongoing costs even at night and at weekends. In addition, an error in the node scheduler prevented the necessary scaling down - at times, more than 30 Kubernetes clusters were running at maximum capacity, generating considerable costs and complexity.
Our solution in five building blocks

1. Cost transparency and FinOps anchoring
We established technical and organizational structures so that all costs could be clearly allocated to the teams. A central architecture and platform team took over FinOps and established regular cost reviews, knowledge sharing and recommendations to the teams.
Added value: clear responsibilities, comparability and continuous optimization instead of ad hoc savings.
2. Align cloud goals with corporate goals
We harmonized cloud usage, operating models and billing logic with the top corporate goals. Incentives from external service providers were made transparent. Result: Investments were made in measures that actually support corporate goals - and not just individual cost items.
Added value: economic decisions instead of technology optimization without business benefits.
3. Cost-benefit decisions in system design
- Dev environment reduced from 3 to 1 availability zone - 6,000 euros less per month, without limiting product goals.
- Principle (cost efficiency > marginal test benefits in stage 1) rolled out to 15 teams.
Added value: Clear guidelines for cost-efficient architecture decisions.
4. Control operating times according to use
- Time-controlled shutdown of dev workloads, from 24 hours to 10 hours on weekdays, 0 hours on weekends.
- Implementation with open source (e.g. Kubedownscaler, KEDA) for elastic workloads.
Added value: More than 4,000 euros less per month - amortized in less than a month.
5. Detect anomalies early - minimize risks
- Monitoring and observability standardized, alerts based on normal cost profiles.
- Scheduler bug discovered through anomaly analysis, 70,000 euros per month in additional costs identified.
Added value: rapid intervention instead of silent cost drivers.
How we proceeded
The project introduced a governance-based approach for the cloud platform, which established a common framework for logging, monitoring, tagging and cost management. This approach ensured that central processes and guidelines were transparent and uniformly accessible for all teams. In addition, a cross-team guild was created as an exchange format in which best practices were systematically shared and scaled as good practices, promoting collaboration and learning throughout the company.
In line with our principle of enablement first, great importance was attached to empowering all teams in a targeted manner so that personal responsibility and transparency in dealing with the cloud platform could go hand in hand. The optimization was carried out iteratively in small, measurable steps that were deliberately rolled out on a role-based basis instead of relying on a comprehensive big-bang approach. In this way, changes could be anchored in the company efficiently, comprehensibly and sustainably.
Technologies & Services
-
AWS (separate accounts per environment)
-
Kubernetes (Autoscaling Groups/Nodes)
-
KEDA
-
Cubedownscaler
-
Observability/Monitoring
-
FinOps processes
-
Cost & tagging standards
What decision-makers learn
Standardized observability and tagging are the basis.
Communication and enablement beat individual optimization.
Operating times, availability and billing models must be supported company-wide.
Up-to-date metrics and alerts are mandatory.
External reviews prevent operational blindness.

Do you want to reduce cloud costs, minimize operational risks and empower teams in your company?
We will show you how transparency + FinOps can make an impact in just a few weeks - and accompany the implementation step by step.
Our references & projects
A reference is worth more than 1,000 words. Fortunately, we have dozens of them. Click through a selection of our most exciting projects and see for yourself!
-
inCTRL Solutions: Modernization of the IoT platform for water treatment plantsTo inCTRL reference
We are making the opsCTRL IoT platform highly resilient, maintainable, and capable of further development. New functions can be integrated into products more quickly.
inCTRL Solutions: Modernization of the IoT platform for water treatment plantsTo inCTRL referenceRenovationof the IoT platform
Maintenance & new featuresgo hand in hand
Change code quicklythanks to CI/CD pipelines
-
TÜV NORD: IT system for damage assessmentsTo the TÜV Nord reference
With the cloud-based system, TÜV NORD experts can efficiently prepare and invoice damage assessments and vehicle valuations. Car dealerships can order spare parts directly after the damage calculation, which shortens the repair time by at least 2 working days.
TÜV NORD: IT system for damage assessmentsTo the TÜV Nord referenceSince August 2023in productive operation
2 daysTime savings for garages
Operation of the systemand the virtual infrastructure -
Research: AI-supported robotics for employees with physical limitationsSee robotics reference
With demographic change, employees with physical limitations need better support to remain productive. Assistance robots can help, but are not flexible enough. The KiRo4LeMi research project aims to use AI to dynamically adapt robots to individual changes in performance. Using digital models and "living personas", the AI optimizes robot operation in real time.
Research: AI-supported robotics for employees with physical limitationsSee robotics reference3 yearsresearch project, funded by the Bavarian Ministry of Economic Affairs
Design of the robotmore individuality thanks to AI
Digital twinhelps to be able to react quickly during operation
-
MAN: Efficient threat analysis for control unitsSee MAN reference
Digitalization increases cyber risks - especially for MAN's new CM4 control unit. Our experts use the 4×6 methodology and ThreatSea to identify threats at an early stage and develop targeted protective measures. Find out how MAN uses intelligent risk analysis to strengthen the security of its vehicles.
MAN: Efficient threat analysis for control unitsSee MAN reference7 monthsProject duration
Over 20 workshopsThreat analyses
Over 500Potential risks evaluated
-
Siemens: AI demand prediction platform for industrial production planningSee Siemens reference
Siemens is looking to the future with the AI Demand Prediction Platform. Thanks to machine learning and AutoML, precise demand forecasts can be created for over 100 products and production can be better planned. Launched as a proof of concept, the platform quickly developed into a system that can be used productively. The self-service web application will soon be used in other plants.
Siemens: AI demand prediction platform for industrial production planningSee Siemens referenceProject durationSince February 2022
Proof of conceptin a few weeks
Time series predictionfor 100 different products
-
TÜV NORD GPT: Development of AI assistanceSee TÜV NORD reference
Find relevant information faster by chatting with documents? It's possible! The TÜV NORD GROUP uses GPT technology in the secure Microsoft Azure Cloud. With the aim of optimizing knowledge management and efficiency. The system enables new usage options within the testing group and is operated securely. Find out more about the innovative AI assistance system now.
TÜV NORD GPT: Development of AI assistanceSee TÜV NORD referenceProject durationsince September 2023
33.000GPT applications in the first month
ChatGPT Model 4in the European Microsoft Azure Cloud
-
Schöck components: Improvement of the requirements processSee Schöck Bauteile reference
Together with MaibornWolff, the construction industry specialist optimized the requirements management of its Scalix design software. Thanks to the digital design approach, the technical requirements were re-specified and the software was further developed in a user-centered manner. The aim is to continuously increase user satisfaction and integrate internal and external interests. Read more about the methodology and results now!
Schöck components: Improvement of the requirements processSee Schöck Bauteile referenceProject durationsince 2022
Team size2-3 Digital Designer:inside
In 2024Scalix replaces old software -
VW: Digitization of key production figures with the iProcess appSee VW reference
Less paper, more efficiency: Volkswagen replaces analog processes with the iProcess app. Cloud-native technologies and digital design enable intuitive recording and analysis of key production figures. This increases transparency and overall equipment effectiveness and paves the way for predictive maintenance at the automotive group.
VW: Digitization of key production figures with the iProcess appSee VW referenceProject durationSince January 2021
Team size5 to 10 persons
Fast app developmentthanks to Digital Design
-
digikoo GmbH: Apple Vision Pro for city plannersSee digikoo reference
Planning cities smarter: Together with TUM, we developed the 3Digipad for Apple Vision Pro. It visualizes complex energy data in 3D and makes scenarios intuitively tangible. Dynamic building data and KPIs support urban planners in making sustainable decisions.
digikoo GmbH: Apple Vision Pro for city plannersSee digikoo reference4 monthsProject duration
3D map with KPIsDisplay building data dynamicallyEnergy dataImmersive visualization -
DER Touristik: Become a digital travel companion in 7 monthsSee DER Touristik reference
Travel information, bookings, support - all in one app. With Flutter, MaibornWolff developed a cross-platform solution for DER Touristik. In just seven months, a stable app was created for iOS and Android that supports multiple brands, languages and countries. This is how customer proximity travels.
DER Touristik: Become a digital travel companion in 7 monthsSee DER Touristik reference7 monthsFrom Kickoff to Go Live
iOS and AndroidDigital travel companion
Whitelabelling solutionUncomplicated integration of additional brands, languages and countries
-
Mixed reality: driving a real car in a virtual worldSee BMW reference
Virtual racing in a real car? MaibornWolff made it possible for the BMW M. Within three months, we used Unreal Engine 4 to develop a prototype that processes vehicle signals with high performance and creates an immersive driving experience without motion sickness. The cockpit remains real, the surroundings become the racetrack.
Mixed reality: driving a real car in a virtual worldSee BMW reference<3 monthsFrom the vision to the prototype
3D applicationsin real time
TeamUnreal development, XR technology, game design, smart devices, 3D content
-
KUKA: Web interface for a new Human Machine InterfaceSee KUKA reference
Rethinking robotics: KUKA is working with us to develop iiQKA.OS - an operating system with an intuitive web HMI. Using web technologies and scrum methods, we enable flexible customization and simple control. Open collaboration makes robotics more accessible than ever before.
KUKA: Web interface for a new Human Machine InterfaceSee KUKA referenceGoalReplacing the current control software
iiQKA.OSFast and intuitive development
> 11,000 hoursProject work
-
STIHL: Control iMOW robotic mower via appRead more
Robot mower redefined: The iMOW from STIHL makes lawn care smarter. Our software expertise combined with STIHL's hardware know-how ensures intuitive app control. The result? A networked device that makes gardening easier - simple, efficient, innovative.
STIHL: Control iMOW robotic mower via appRead moreProject durationsince August 2020
Robot mowerControl via app
Process optimizationwith the customer
-
DEKRA: Modern enterprise architecture thanks to co-creationSee DEKRA reference
New IT for a sustainable future: the global auditing group needed to modernize its IT landscape. Co-Creation created a harmonized architecture and the EA Community promotes cross-border exchange for strong IT until 2025 - and beyond.
DEKRA: Modern enterprise architecture thanks to co-creationSee DEKRA referenceMost important method:Building an EA community
Created conceptEnterprise Architecture
Cooperationat eye level
-
Miele domestic appliances are networked worldwideSee Miele reference
How do household appliances become smart? Miele relies on an IoT platform that we played a key role in developing. Container-based architecture ensures stable communication, DevOps for continuous improvement - so that smart appliances can be found in millions of households worldwide.
Miele domestic appliances are networked worldwideSee Miele referenceProject durationsince 2016
Success factors:High availability and scalability
IoT platformwith container-based architecture
-
MAN: Secure Software Development Life CycleSee MAN reference
Trucks are networked - and therefore unfortunately also targets for attack. With SSDLC, MAN relies on security by design and integrates the highest security standards directly into development. Together with 22 teams, we have established security measures that go beyond the legal requirements.
MAN: Secure Software Development Life CycleSee MAN reference22 monthsProject duration
UNECE R155successfully integratedMethodsSecurity Champions, DevSecOps, OWASP SAMM
-
ifm services: Remote maintenance of systems and machinesSee ifm services reference
Maintenance, anywhere and at any time - together with ifm services, we developed a remote access solution for industrial plants. A small, agile team created a full-stack cloud application that combines intuitive operation and secure connectivity. The product celebrated its premiere at HMI 2024.
ifm services: Remote maintenance of systems and machinesSee ifm services referenceProject durationSince March 2023
Remote Accessintegrated in platform
Team5 Developers
-
It's LOS: Building a cloud-based low-code applicationSee Es geht LOS reference
Democracy meets digitalization: For "Es geht LOS!", we developed a low-code app for candidate management in five weeks. Built on AWS, it enables secure draws and efficient user management - for more citizen participation digitally.
It's LOS: Building a cloud-based low-code applicationSee Es geht LOS referenceStrengthening democracythrough random selection
Recruit candidatesfor citizens' councils via app
In 5 monthsfrom concept to prototype with low code
-
Monitoring alarms in industrial plantsSee reference
Safety in industrial plants requires a real-time overview. We developed a live monitoring platform that networks measuring devices and collects data via Azure IoT and Kubernetes. A scalable MVP was created in just three months: alarms appear in under ten seconds thanks to automated data acquisition via smartphone and Bluetooth.
Monitoring alarms in industrial plantsSee reference3 yearsProject duration
<3 monthsto the MVP
Cloud migrationModernization & relocation -
Planning systems: Optimizing the capacity utilization of pressing plantsSee reference
Maximum capacity utilization, minimum costs: Our customer relies on a central planning system for pressing plants. We have been supporting its further development with .NET Core, GraphQL and PostgreSQL for over ten years. Now the journey is moving to the cloud - for greater scalability, flexibility and efficiency.
Planning systems: Optimizing the capacity utilization of pressing plantsSee reference>10 yearsCustomer support
Technological modernizationCloud-ready platformEfficiency through dataSite-specific cost planning -
Global workforce planning systemSee reference
How do you achieve globally harmonized workforce planning? With a modern web application based on domain-driven design and hexagonal architecture. Event sourcing ensures transparent traceability, while Azure and Quarkus provide stability and scalability. The result: a flexible, future-proof solution.
Global workforce planning systemSee referenceTeam sizeUp to 8 employees
Domain Driven DesignDevelopment approach
Microsoft Azure Cloudallows easy roll-out of new versions
-
DER Touristik Online: Development and migration of a multi-client capable travel booking platformSee DER Touristik reference
Different brands, one platform: DER Touristik merges several independent websites. We are supporting the migration to the AWS cloud with digital design, microservices and testing. The focus is on performance, SEO and UX - for a seamless digital travel experience.
DER Touristik Online: Development and migration of a multi-client capable travel booking platformSee DER Touristik referenceStandardized platformMulti-client travel portalQuality assuranceEnd-2-End test automationCloud transformationMigration to AWS cloud -
KUKA: UI/UX design for an app for load data analysis for industrial robotsSee KUKA reference
How do you reduce support requests? With a smart UX! KUKA and MaibornWolff developed a web-based application for load data analysis. Through "Understand/Build/Learn" we recognized challenges early on and validated solutions for an intuitive user experience.
KUKA: UI/UX design for an app for load data analysis for industrial robotsSee KUKA referenceMethodProduct Experience Design
Customer requestEasier interaction between users and the system
Our goalOptimized load data analysis, fewer support requests, higher satisfaction -
BMW Group: Replacement of a production-critical legacy systemSee BMW Group reference
20-year-old system, new digital future. We gradually migrated production software to a flexible architecture. After two years, central plants were running on the new solution - stable, fail-safe and ready for the challenges of tomorrow.
BMW Group: Replacement of a production-critical legacy systemSee BMW Group referenceProject durationMid-2018 to the beginning of 2024
Team size25 employees
Legacy systemreplaced after 20 years
-
BMW Group: Virtual reality brings vehicle design to lifeSee BMW Group reference
From the first draft to series production - BMW uses VR to visualize the entire development process. Based on the Unreal Engine, a uniform platform visualizes designs and processes. This saves on hardware prototypes and enables location-independent collaboration. MaibornWolff plays a key role in driving the development and integration of these VR use cases.
BMW Group: Virtual reality brings vehicle design to lifeSee BMW Group reference3.2 yearsProject duration
CooperationLocation-independent and virtual
InfrastructureScalable and expandable thanks to the cloud
-
BMW Group: Remote software upgrade for vehiclesSee BMW Group reference
No visits to the workshop, no stress - thanks to Remote Software Upgrade, BMW vehicles worldwide remain up-to-date "over-the-air". MaibornWolff has been developing and operating the backend for secure updates for over three years. Microservice architectures and DevOps approaches guarantee stability, performance and security.
BMW Group: Remote software upgrade for vehiclesSee BMW Group reference5 yearsProject duration
Millions of vehiclesreceive new features thanks to "over-the-air" upgrade
IT securityAlways up-to-date thanks to remote software upgrade
-
digikoo: A data platform for the Azure CloudSee digikoo reference
Structured geodata, automated quality assurance, seamless provision - we developed a powerful Snowflake data platform on Azure for Digikoo. It makes analysis easier for data scientists and lays the foundation for precise forecasts and new use cases.
digikoo: A data platform for the Azure CloudSee digikoo reference5 monthsProject duration
Climate changePlan digitally and implement efficiently
Foundation data platformMicrosoft Azure Cloud
-
Creditreform: Secure proof of identity on the webSee Creditreform reference
Verifying online identities in a forgery-proof manner while maintaining user-friendliness - CrefoTrust makes it possible. Together, we developed a solution that creates trust with blockchain technology and proofs of concept. Personal and company data remain protected, while verifications run smoothly.
Creditreform: Secure proof of identity on the webSee Creditreform reference5 yearsProject duration
Pilot testsbased on several PoCs
Decentralized identityallows tamper-proof storage of identities and company information
-
Bayernwerk: Knowledge management via teamsRead more
Experience must not be lost - Bayernwerk digitizes the knowledge of long-standing employees. MaibornWolff designed an intuitive MS Teams app with a clear UX/UI. Close collaboration, lived Scrum values and user-centered development make the app a success. Promoting exchange, optimizing processes - this is how knowledge transfer works today.
Bayernwerk: Knowledge management via teamsRead more6 monthsProject duration
Target:Identify implicit knowledge
Requirements:a user-centered, intuitive and clear UX/UI design
-
Weidmüller: Progression of the Industrial Service PlatformRead more
IoT meets agility - Weidmüller and MaibornWolff developed the easyConnect platform for remote access, data visualization and machine learning. After intensive exploration, an MVP was created, accompanied by a cross-functional Scrum team. Making Industrial IoT smarter and more efficient.
Weidmüller: Progression of the Industrial Service PlatformRead more12 monthsup to the MVP
8 weeks analysisof professional, technical and organizational factors
Innovative portalfor end-to-end solutions
-
SMA: Development of a Web UI for ennexOS platformSee SMA reference
Optimizing energy flows, reducing costs - we have been developing the Web UI of the ennexOS platform with SMA Solar Technology since 2016. It digitizes energy management processes for over a million users and combines smart solutions for a sustainable energy future.
SMA: Development of a Web UI for ennexOS platformSee SMA referenceProject durationsince 2016
> 1.000Modules and components in ennexOS platform
> 7.000Fuse tests
-
Travel information systems: 25 percent savings in cloud costs and stable operation thanks to FinOpsTo the FinOps reference
Cloud transparency and FinOps governance for a networked travel information system - with clearly measurable savings and less operational risk.
Travel information systems: 25 percent savings in cloud costs and stable operation thanks to FinOpsTo the FinOps referenceCloud operating costsreduced by 25%
Reductionof the availability zones in the development environment
Time-controlled switch-offof the development environment
-
Supply chain management: Reducing cloud operating costs by 50 percent with FinOpsTo the FinOps reference
We have made the supply chain management system for an international industrial company more scalable and robust through process modernization, improved monitoring, automation and rightsizing. Targeted optimization of the infrastructure reduced oversizing and unnecessary resources - the platform is more stable.
Supply chain management: Reducing cloud operating costs by 50 percent with FinOpsTo the FinOps referenceLoweringof the cloud operating costs
Fastto the new release
More transparencyand control
Why MaibornWolff?
As one of the most innovative IT service providers with a great passion for AI, we focus entirely on the project business and individual software development - without our own products. To stay at the forefront, we continuously invest in our team of digital technology engineers and develop digital solutions that are well thought-out, efficient and reduced to the essentials.
Our principle: simplicity instead of complexity. We only develop what is really needed - tailor-made, useful and reliable. Our results speak for themselves. With over 800 large-scale systems and more than 10,000 person-years of experience in high-end software engineering, we are one of the few who can reliably implement even the largest and most complex IT landscapes. Thanks to close partnerships with leading hyperscalers, our customers operate their solutions in today's most modern and powerful environments.
Less technology. Better Business.