Now: Scalability in Data & AI is a big topic
Companies today face many questions about scalability in Data & AI and are looking for answers. They are aware that a non-scalable system will eventually reach resource limits. This must be prevented at an early stage. Dr. Carsten Bange, Co-Founder of DATA festival and CEO of BARC, believes “Any plan to become more competitive using data & AI should also continue to be persuaded.”
Companies like the Lufthansa Group or Scout24 AG have been successfully tackling this challenge. At the DATA festival #online 2022, they shared their experiences with attendees and made clear: the best-practice examples and lessons-learned of a scalable system exist already in companies.
Good AI isn’t just accurate, it is transparent and scalable
Xavier Lagardere, Chief Data Officer at Lufthansa Group brought a vivid example of his company. For him and his team, it is important to advance data analysis at Lufthansa to get to know the customer even better and thus offer better quality. “Every 24 hours, when the new aircraft of Lufthansa flies, it generates 1 TB of data. And it’s just operational data! All this data should be analyzed in order to use it as insights for a better understanding of the customer. Their goal: to become a data-driven airline for all core business processes.
A roadmap with seven key data initiatives, should support them in the transformation and give an overview. This is basically about the 4 blocks: Data Roadmap, Data Quality & Governance, Data Community & Transformation, and Data Science Delivery.
In his presentation, Xavier also talks about how important it is to build a fundamental data culture in the company, as this allows data transfer to take place more effectively.
As a final point, he emphasizes: “It’s a human transformation. Any data-driven transformation starts and ends with enabling the human beings (operators) to work with the recommendations.”
That’s why his team and him need to give employees enough confidence in the systems so they can see the sense behind them. “By no means, we are looking at AI and automatization as a way to replace humans but actually to simply manage our operations, by bringing automation to humans.”
“The purpose of a data platform within a company is to scale the work with data”
Arvid Reiche, Data Platform Leader at Scout24 AG, is also in the process of creating scalability for his company. He and his team have been building a scalable data platform since 2014.
“The purpose of a data platform within a company is to scale the work with data.” The overall goals for this project were set from the beginning: Get rid of all data warehouses, empower data users and data producers, create a data culture, and reduce the cost of data technology.
Among many milestones that Arvid has achieved with his team over the years, there are some lessons-learned that he shared in his talk. “In the past, we had a wild western pattern which means that we empowered everyone to do everything, but there were no consequences. (…) We were stuck in a data swamp instead of swimming in a crystal-clear, plastic-free-data ocean.”
As a result, the three major cornerstones of Data Governance, Data Culture, and Data Mesh have emerged at Scout24.
This transformation is also very much about empowering employees and reducing unnecessary costs and confusion. For the next few years, he and his team will continue to work on their big vision: “We transform Scout24 into a data-driven company to help make the experience of our users simple, efficient, and stress-free.”
The complete presentations by Xavier Lagardere and Arvid Reiche are available free of charge on the On Demand platform of the DATA festival #online 2022.
Check out the talks about “Scalability in Data & AI“ here:
AI at scale: Becoming a data-fueled airline across core business processes
(Speaker: Xavier Lagardere, Chief Data Officer @ Lufthansa Group)
Day 1, 27th october 09:00 – 09:30
- What is this talk about?
The data transformation journey of the largest European airline group: priorities and key opportunities & challenges on the way to becoming a true data-centric organization.
Scout24 – Building a scalable Data Platform: lessons-learned by seven years Data Manifesto
(Speaker: Arvid Reiche, Data Platform Leader @ Scout24 AG)
Day 1, 27th october 11:00 – 11:30
- What is this talk about?
Ever since ImmoScout was founded, we are crossing technical frontiers. When we migrated fully to the cloud in 2015 and 2016, we were the pioneers. When data engineering infrastructure became famous, we already had years of experience building a data platform and when Data Mesh became popular, we already had something similar in place for almost five years. Only by introducing new terms like data democratization, data culture, etc. data platforms do not scale. We are happy to share our learnings, failures and success of building a custom data platform since 2015 with you.