With SAPPHIRE 2024 behind us, I here take the opportunity to collect my impressions and highlights of the event. Even if subtle, this year marks a very important milestone for SAP, its very first “Corporate GenAI positioning and strategy”. Sure GenAI Hub has been there for a little while but 2024 is the year where things start to materialize, really. The GenAI train (industry) rolls at 200Km/h, will SAP be able to act fast? It seems SAP is on the right direction but it won’t be easy and possibly some clarification and focus is required. I will comment along the list of highlights.
- SAP has announced that Joule will be available across SAP Cloud products.
My take:
This is a coherent approach as we need a single user experience across multiple SAP solutions. Also, some capabilities should be cross-system. So I believe this is a good move.
Also, it reinforces the strategy of SAP of driving innovation in the cloud. Rise and Grow are clearly center to this discussion.
We continue to follow SAP’s pace on delivering on that promise. Joule was announced last year and it is becoming a reality now. The pace of innovation around GenAI is very fast. Will SAP be able to follow that pace? I think it can, with the help of partners. Joule needs to be open. Open for extensions and open to consume services from other platforms. (more on that later)
- SAP has announced that Joule be delivered with 2 new capabilities, Joule for Consultants and Joule for ABAP Developers.
The first, Joule for Consultants, will be one that has grounding (knowledge base) around help.sap.com, training documentation and a lot of more SAP related content. So this is quite promising. Don’t remember seeing a mention on availability.
The Joule for ABAP Developers however is one that needs more ground work from SAP as it requires a fine-tuned LLM. I’ve heard it is scheduled to be delivered in Q4’24 (so we should expect access in Q1 2025)
My take:
These are amongst the best announcements for me at SAPPHIRE. The industry is rely needing this. As we work in this field everyday, it is clear that there are areas that we need SAP to drive. That is:
I go even one step forward, I believe SAP should make these groundings available (by subscription if needed be) so we can build our GenAI solutions leveraging them. A lot of players in the industry are doing exactly that and it makes no sense to reinvent the wheel.
- SAP is enhancing its Business AI capabilities, providing users with advanced tools and insights to improve outcomes and decision-making.
My take:
Two important moves from SAP have been made in recent years, and that will make Business AI adoption better.
1 – Intelligent Scenario Lifecycle Management is the new APP that consolidates all Business AI innovations inside S/4. It makes easy to browse and navigate.
2 – Pricing around innovations has moved to a AI-UNITS consumption model. As we focus on the midmarket, where volumes are smaller than large enterprise, making business cases have just been made simpler.
- SAP is expanding its collaborations with technology leaders like AWS, Microsoft, Google Cloud, Meta, Mistral AI, and NVIDIA to push the boundaries of what generative AI can achieve at an enterprise level.
My take:
This are great news but it will all come down to tangible improvements and innovations. We will need to continue following to make sure when this will be made available. Some are more immediate like the LLM models of Meta and Mistral AI. Nvidia’s partnership is a promising one but we need to wait until it is delivered (referring here to Joule’s capabilities and Joule for Developers.
- The SAP Business Technology Platform is incorporating large language models from AWS, Meta, and Mistral AI into its generative AI hub.
- This integration simplifies the development of generative AI use cases for SAP applications, facilitating seamless access to information and improved automation.
My take:
GenAI Hub is needed. We need SAP to deliver on this. We need to model and build GenAI capabilities in BTP development. I saw the early preview of the new release and it is very promising. It will deliver guardrails capabilities, Grounding management, more LLM availability, etc
But while I believe SAP needs to play in this field, providing technical tooling and capabilities, it shouldn’t be done with a goal of delivering the best (and sole) genAI platform around SAP. There are so many alternatives out there that it will be impossible for SAP to lead (and I honestly think it should not even try to). The ideal situation is one that this platform is open. It needs to be able to consume features from other platforms and be able to be consumed by other platforms. It needs to be able to handle local LLMs and consume already purchased hybrid LLMs (many companies have already contracts with Azure for OpenAI, same for other hyperscalers like AWS)
Where I am confident SAP can win is on its data. Let me be more precise:
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
3 | |
3 | |
2 | |
2 | |
2 | |
2 | |
1 | |
1 | |
1 |