Success in 90 days with AWS and MHP

How scalable, secure and future-oriented cloud applications are created

In many companies, IT projects fail due to unrealistic timelines, a lack of coordination or an unclear target vision. The challenges in IT security and compliance require careful consideration, as failure to do so can jeopardize the go-live of already validated concepts in productive use - the “proof-of-concept trap”.

Despite the high demands, the project was completed successfully and on time. This was due in part to a plan developed by MHP and the efficient use of the Well-Architected Framework from AWS, which offers best practice guidelines for secure and high-performance cloud architectures. The active transfer of knowledge, which MHP supported significantly, was also crucial. This enabled the knowledge gained to be sustainably integrated into the team at Porsche and made further initiatives possible.
The initial situation was a Python solution operated on local end devices for the detailed analysis and optimization of procurement processes. Employees were manually comparing a large number of Excel spreadsheets. Data points from different files had to be laboriously transferred into new documents. This was prone to errors and lacked transparency. This meant that data-driven decisions could not be made quickly. Porsche's goal was therefore to simplify processes with the help of the cloud. 

The Well-Architected Framework from AWS offered the right approach: first, the analysis process in procurement was significantly simplified in a PoC implementation. The code could be transferred from the Excel tables to a Python-based solution and the initial results of the new approach were quickly convincing.

The difficulty now was to make the concept available to all colleagues, as the application could only be run on a single laptop. The lack of central hosting, high manual effort and a significant risk of errors characterized the picture. The brief was to make the PoC fit for productive use in 90 days as part of a consistent shift to the cloud - in such a way that it scales as a highly available, container-based application, runs securely and enables future AI applications. All with the aim of optimizing the work of the procurer in the analysis and optimization of procurement processes, achieving better negotiation results and increasing efficiencies.

From laptop script to cloud revolution: how to make the shift to the cloud in record time

Adrian Grünther: ”In addition to rapid implementation, our aim was to coordinate the sound methodology with an innovative technology and the close integration of business and IT.”

The solution had to meet all the requirements of the six central pillars: Operational Excellence, Security, Reliability, Performance, Cost Efficiency and Sustainability. MHP used the framework as a compass to keep everyone involved in the project on the same principles and prevent important details from getting lost in the tight timeline.

Security-by-design according to the well-architected approach

Security was the first priority in order to make every Excel file upload process absolutely reliable. A sophisticated role and authorization model determined who has access to the data and how every action is seamlessly recorded. Security by design according to the well-architected approach also meant that all data was encrypted, any vulnerabilities were systematically identified and appropriate countermeasures were defined.

At the same time, the project team ensured that requirement and configuration changes were integrated in a controlled and orderly manner during the shift to the cloud. By consistently embedding DevOps practices in the well-architected approach, every code change was quickly tested in an automated pipeline and rolled out in the new environments. This continuous deployment process ensured the required operational excellence, as errors were detected and rectified immediately before they could affect productive operations.

Demand-oriented scaling

At the same time, the team attached great importance to reliability and performance efficiency. With the help of a container-based architecture, it was possible to scale the Python application as required. For example, if the number of users increases, additional containers are automatically provided. If the workload decreases, the amount of resources required is reduced again. This dynamic concept, based on the Well-Architected Framework, prevents expensive overcapacity and guarantees high availability even during peak loads. The whole process was supported by structured monitoring via cloud monitoring and alarms in the event of bottlenecks so that the project team could intervene in good time at any time

Cost-effective and demand-oriented scaling

The aspect of cost efficiency was also of crucial importance. Thanks to the pay-per-use model in cloud operation, only the resources actually used are paid for. Combined with the ability to scale automatically, this ensured that the system scales cost-effectively under high load and does not generate any unnecessary additional costs during periods of lower utilization. This focus on cost efficiency is also part of the Well-Architected framework and shaped not only the technical architecture, but also the way the project planning was set up.

Agile product team in close collaboration

Under these conditions, the team successfully adapted to the new requirements and continuously obtained feedback from the procurers. The traditional boundaries between business and IT were deliberately removed, which led to close and effective collaboration. Despite the high time pressure, both user acceptance and technical stability increased.

Knowledge-Transfer & Enablement

The team at Porsche was closely involved in all phases to ensure effective team enablement. In addition to close coordination in the planning of user stories, active knowledge management was also carried out within the team at a technical level. Two main approaches were used for this: Firstly, we used pair programming, where team members work together on code and share their knowledge directly. Secondly, code reviews were implemented to ensure that everyone benefits from the knowledge of their colleagues. Regular workshops and feedback sessions promoted the exchange of best practices.

Project objectives fully met: secure, scalable, efficient and cost-optimized

Bernhard Rack, Product Manager in the area Controlling & Unternehmensreporting at Porsche: “Our collaborative journey - making the leap from a localized proof of concept to a highly available, AI-enabled cloud application - is thanks to the knowledge transfer that MHP actively supported. The knowledge transfer helps us to carry insights and learnings further into the organization, which has a positive impact on parts of our transformation.“

After exactly 90 days, the cloud solution successfully went live. Where previously a locally operated Python application was in operation on a single laptop, a modern, containerized platform is now available. The new solution processes uploaded Excel files in procurement in the shortest possible time, scales automatically as required and has several advantages over the manual approach: Speed, availability, security and future viability.

From a business perspective, the project impressively demonstrates how a scalable, secure and future-oriented cloud application can be created from a locally limited proof of concept. Instead of just one employee, the analyses are now available to all authorized procurers, which significantly increases the benefits for the company. Thanks to the automated, container-based processing of Excel files in the cloud, Porsche is more flexible and faster in the analysis and optimization of procurement processes compared to before, which has a positive effect on the preparation of negotiations.

Capacity bottlenecks are also eliminated, as the system can easily absorb peak loads using additional containers. Thanks to the AWS Well-Architected Framework, which covers all six core areas from security to cost efficiency, high security and performance standards are achieved while only paying for the resources actually required. Moving the application to the cloud enables seamless AI integration by providing scalable computing power, access to cloud-native AI services and centralized data for model training and real-time processing. This provides Porsche with a stable platform for driving data-driven processes and tapping into further innovation potential in procurement. The close integration between the specialist department and IT, the iterative approach and the strict adherence to the Well-Architected principles are success factors that can be transferred to many other initiatives. This success story impressively illustrates the potential that can be realized when sound methodology and innovative technology are aligned - a small PoC is transformed into a viable, production-ready solution in a very short space of time, which helps companies like Porsche to make sustainable progress.