Recapping Data to Value Day: Collibra, Google and SAP take center stage together


Every year SAP organizes the Data to Value Day: an event during which it showcases customer cases using the latest of SAP technology, reaffirms its vision for the second half of the year and provides attendees with the perfect opportunity to meet SAP representatives and get their networking game on. This year’s edition was something special however, and not just because of the rapidly evolving AI solutions that dominate the headlines. For SAP’s increasingly expansive partnerships bore fruit, with Collibra and Google joining the day to talk about how they work together with Walldorf to help solve customer problems. In this blog, I will recap the most important takeaways from the day that Collibra, Google and SAP took center stage together.


SAP Datasphere’s (DSP) March update brought three milestones to the table: the new Graph Builder, Google BigQuery integration and AI Governance with Collibra. While the first two items had been on the solution’s roadmap for quite a while already, the latter is most noticeable primarily because of the emphasis SAP has been placing on its Business AI portfolio in the past months. If you look at SAP’s data-based timeline, statistics have already played a vital part in the ecosystem for decades, while Machine Learning only picked up in the early 2000s with HANA’s advanced and predictive analytics (think of the SAP Predictive Analytics tool and the APL and PAL). Generative AI is the next step in this evolution and the focus of today, following up on Deep Learning (e.g. AI Core) with Large Language Models (LLMs), HANA’s Vector engine and Foundation Models being at the core of SAP’s strategy in 2024.

SAP's Business AI is founded on the three R’s of AI: Relevant, Reliable and Responsible (more on these concepts can be read here). Coincidentally, there are also three main components within SAP’s AI approach: Joule, the company’s copilot, embedded AI, which is increasingly available across the SAP ecosystem and finally, the AI Foundation. The latter provides customers with a toolkit on the BTP that allows them to build their own solutions; solutions that can answer needs not met by SAP’s proprietary offerings (for example the new premium document information extraction). These tailored solutions are based on Foundation Models, which in turn are either built by SAP or accessed via hosting or remotely. Customers can also leverage a variety of LLMs from different providers, such as Azure OpenAI, Google Gemini Pro or PaLM 2, which are licensed through a single contract with SAP (which is quite handy). If you want to learn more about SAP’s Business AI, I highly recommend this learning journey.

Content below: An overview of the SAP AI Foundation – Source: SAP

Blog Data to Value Day 1


Google opened their pitch by stating that their partnership with SAP dates back years, and that Alphabet also employs SAP solutions to run their daily operations (all Play Store transactions go through an SAP system, for example). Google’s Document AI and Workspace have also been used to support SAP’s Build Process Automation for about two years already, and the two companies are also cooperating on SAP’s Rise initiative. Now, with a focus on a ‘better together analytics environment’, BigQuery and Vertex AI form the foundation on which SAP solutions close the data to value gap by retaining business context across the platform. This is expressed in the cross-data enrichment between SAP and Google’s AI platforms (see image below) as well as Google Cloud Platform as a value-adding hyperscaler for SAP Datasphere.

Content below: SAP Business AI working together with Google’s Vertex AI - Source: Google.

Blog Data to Value Day 2

SAP Datasphere data can be replicated to BigQuery and vice versa, whereas federation is possible from DSP to Vertex AI and from BigQuery to DSP. Google’s integration with SAP also materializes towards Analytics Cloud (SAC), with BigQuery planned to be the first supported SQL-based live source for SAC in Q4 of this year. Other Google offerings, such as Cortex, can still be used in conjunction with SAP Datasphere to leverage the latter’s data and build specific models. Contemporary use cases for the collaborative effort between SAP and Google that were mentioned included Marketing, Supply Chain Management, Finance and Manufacturing.


Collibra was quick to position itself more as a business-oriented solution as opposed to a technical one. Of course, there is still a lot of smart thinking that goes into realizing the cataloguing, governance and lineage that their solution provides. The current outlook is that all EU-based companies will have to conform to the AI Act in 2025 (in addition to the Data Act, which entered into force this year), which emphasizes the need for reliable, traceable and compliant AI practices. This puts Collibra in an important spot on IT Architecture sheets, with its insistence on effective AI that puts a comprehensive user interface (UX) next to GRC practices as a critical priority. Collibra achieves this by leveraging metadata from the relevant source systems, whether it be SAP, Google, Databricks (integrations that are currently live) or in the near future, SAP’s Business AI Foundation, Azure or AWS.

In Collibra’s vision, there are three main approaches to AI governance: Model-ops centric; which prioritizes model performance and maintenance, compliance centric; focused at sticking to regulations and transparency and finally, data centric; the paradigm that Collibra maintains. This approach entails looking at data as an asset that should be used with purpose, and not just accumulated. In turn, this also means that solutions should be designed around relevant data, instead of the other way around. This goes double for AI, and since AI is non-deterministic, it requires constant monitoring and verification to prevent occurrences of hallucinations or drift. These capabilities are something Collibra provides with their solutions (or can assist a governing body, such as an AI governance council, with).

Content below: An example of an SAP Datasphere diagram view in Collibra (private beta) - Source: Collibra.

Blog Data to Value Day 3

In relation to the SAP ecosystem, the integrations that are available today include SAP HANA, ECC, S/4HANA Public Cloud, BW, SuccessFactors and Sybase ASE. Collibra’s focus is currently on Datasphere and Analytics Cloud (SAC), through which it utilizes special APIs that are accessible only through their software. The demo we were shown utilized the integration of metadata with SAP Datasphere, which had only been launched in private beta the day before. The results were impressive, with Collibra’s data lineage going deeper than SAP Datasphere’s own catalog (which is harvested by Collibra) even when ‘drilling down’ to SAC. Although this specific integration is thus still in the private beta, I see a lot of potential here for SAP customers, especially with Collibra’s focus on AI governance and roadmap in mind (with BW Bridge integration planned for later this year and BW/4 and S/4HANA Private Cloud planned for 2025).

Special thanks to all organizers and speakers of the Data to Value Day 2024: Marlou Houdijk, Niels van der Kam, Marcel de Bruin, Ronald Lanjouw, Wouter Mertens, Anouk Gorris, Steve Wainwright and Victor Palkin.

Blog Data to Value Day 4

Charlotte is here to listen.
Get in touch with her.

%firstName% is here to listen.<br />
Get in touch with %pronouns%.

About the author

Photo of Lars van der Goes
Lars van der Goes

Lars van der Goes is a SAP Data & Analytics Lead at Expertum. Lars combines strong functional skills with broad knowledge of Analytics principles, products and possibilities.

Read more articles by Lars van der Goes