A Quick Recap on SAP TechEd Berlin 2025 Announcements & Insights

After three incredible days in Berlin, endless coffee runs, packed sessions, and even longer evening chats with fellow attendees, I’m finally sitting down to reflect on what stood out most at SAP TechEd 2025.

It was a fantastic event, buzzing with AI-powered innovation and partner-focused announcements. Between deep-dive technical sessions, hands-on workshops, and countless hallway conversations, one thing was clear: SAP’s innovation engine is running at full speed. From AI-driven agents and Clean Core evolution to cloud-native development patterns and open integration architectures, the pace of change is nothing short of remarkable.

For me, one of the biggest highlights was the announcement of VS Code support for ABAP, along with the introduction of SAP-RPT-1 and SAP-ABAP-1 models. These moves really underscore SAP’s push toward more modern, developer-friendly workflows. 

And of course, one of the best parts of TechEd was meeting people I’d only seen on Teams calls. It was a wholesome reminder that our community is not just virtual, it’s alive and thriving.

I am glad that it’s finally back in person, because you can’t replace the brilliance of meeting face-to-face, sharing ideas (and a beer or two).

Day 1: The Big Announcements That Set the Tone

• Free SAP Build TDD Licenses for All Partners

The day started strong with a headline announcement: SAP Build test, demo, and development (TDD) licenses are now free for all SAP partners. This means no more barriers to innovation. 

With free access, partners can prototype, test, and launch AI-powered applications on SAP BTP without initial licensing costs. Access is available with entitlements up to “fair use limits” and can be managed through the SAP Partner Portal. (Read the full SAP announcement.)

The free TDD access includes:

• All SAP Build capabilities (low-code, pro-code, and generative AI tools)
• SAP HANA Cloud
• Additional services within the Generative AI Hub

For many of us, this felt like SAP saying: “Go ahead, build something amazing, and we’ll take care of the entry cost.”

It’s a bold, community-driven move that encourages hands-on exploration and faster value creation for partners. 

Suggested Reading:

How to Become an SAP Build Partner to Gain Access to Licenses

• Prompt Optimiser in SAP AI Foundation

If you’ve ever worked with multiple LLMs, you know the pain. Switching between models like GPT-4, Claude, or Gemini can feel like starting from scratch every time. Each one interprets prompts a little differently, and fine-tuning them manually is a slow, trial-and-error process.

SAP just made that smoother. The new Prompt Optimiser, integrated into AI Foundation on BTP, automatically fine-tunes prompts across models so you don’t have to keep rewriting them yourself.

We’re talking about some serious efficiency gains here.

• No more manual re-engineering of prompts
• Accuracy improvements between 5 % and 60 %
• Tasks that used to take 4 hours can now be done in under one hour.

It’s one of those behind-the-scenes tools that can genuinely transform productivity for consultants experimenting with multi-model AI workflows.

For anyone building generative AI scenarios on BTP, this tool is going to save time, frustration, and plenty of late-night debugging sessions.

• SAP Rapid1 (RPT-1): A Foundation Model for Business Data

Perhaps the most technically exciting reveal that got every developer in the room leaning forward: SAP Rapid1 (Relational Pre-Trained Transformer – RPT-1), the first foundation model built for structured business data.

There are three versions:

• RPT-1 Small – super-fast, tuned for real-time responsiveness
• RPT-1 Large – extra-accurate, meant for deeper analytics
• RPT-1 OSS – open-source and available right now for anyone to explore

What makes it stand out is how practical it is. RPT-1 can deliver predictions directly on your business data without the usual ML lifecycle overhead: no model training, no deployment pipelines, no GPU-burning compute sessions. 

You just connect your data and start generating insights. And the stats are wild:

• Up to 3.5× better prediction quality than comparable models
• 50 000× less energy usage and 100 000× fewer GPU FLOPs
• Runs 50× faster, with a 20 % price reduction across all models
  

For developers, this is a breath of fresh air, something truly optimised for tabular, structured enterprise data, not just text. 

You can test it through the Generative AI Hub, alongside models from Anthropic, OpenAI, Gemini, and Mistral, all under a 30-day free trial.

As someone who’s spent more nights than I’d like to admit wrangling training scripts and model deployments, seeing this level of performance with such a lightweight footprint feels like SAP just rewrote the AI playbook for business data.

• SAP × Snowflake Collaboration

Yes, you read that right. SAP announced a partnership with Snowflake. They’re calling it SAP Snowflake, and it’s a move that finally bridges the power of Snowflake’s data cloud with SAP Business Data Cloud (BDC).

For anyone who’s ever had to juggle data between SAP and non-SAP systems, this is huge. It means we’ll be able to connect existing Snowflake environments directly to SAP BDC through BDC Connect, enabling zero-copy data sharing. In simple terms: no more exporting, transforming, and duplicating data just to get insights across systems.

With this integration, you’ll be able to work with real-time, semantically rich business data (SAP and non-SAP) within the same ecosystem. 

That’s a major step forward for companies running hybrid data landscapes.

SAP Snowflake is slated for Q1 2026, and BDC Connect for Snowflake is expected in H1 2026. 

For me, this announcement represents more than just another integration; it’s a clear signal that SAP is fully embracing an open, connected data ecosystem. It’s not only acknowledging the tools many of us already use, like Snowflake, Databricks, and Google Cloud, but actively building bridges to make them work together. That’s the kind of openness the enterprise world has been waiting for.

• 30-Day Free Trial of AI Foundation

Another great move from SAP this year was the 30-day free trial for the AI Foundation on BTP. For anyone curious about building AI-driven scenarios or just wanting to get their hands dirty with the new generative AI tools, this is the perfect opportunity.

I see this as SAP’s way of saying, “Go ahead, explore, break things, learn fast.”

The trial gives developers, partners, and customers full access to the Generative AI Hub, which includes models from OpenAI, Anthropic, Gemini, and Mistral. You can build, test, and refine ideas directly on BTP without any financial risk.

I love this because it lowers the barrier to innovation. Instead of long approval cycles or waiting for sandbox access, teams can just try things out, and that’s often where the best ideas are born.

• SAP Certification Reimagined

Another thing I really appreciated at TechEd Berlin this year was the renewed focus on learning and certification. SAP reimagined its certification model as open-book, scenario-based assessments that test what you can do, not just what you can memorise.

Many developers got certified right at the event, for free, in areas like SAP Build Development or Generative AI. I saw so many people leaving TechEd with shiny new credentials and even bigger smiles.

To help you prepare, SAP has made its learning ecosystem more accessible than ever:

• learning.sap.com now offers over 100 free learning journeys focused on BTP, AI, and SAP Build
• The SAP Discovery Center is your go-to for reference architectures, best practices, and real customer stories
• And the BTP Guidance Framework brings together decision guides and methodologies in one place

Session Highlights & Key Takeaways

TechEd wasn’t just about big announcements; there were technical sessions, community talks, and hands-on workshops to truly sense the collective direction SAP is heading in. I came away with a few notes from some of my favourite sessions.

• BTP as the Foundation for Partner SaaS Solutions by Gert-Jan

Gert’s community talk was packed with practical wisdom. He ran an insightful session on how partners can leverage SAP BTP as the backbone for SaaS innovation.

He shared a lot of practical advice that hit home for anyone who’s ever built (or struggled to maintain) SaaS extensions on SAP:

• Start with best practices. CAP for backend logic, Fiori Elements for UI.
• “Pluginify” your solutions to reduce maintenance effort over time.
• Ensure robust logging and monitoring from day one; debugging in production isn’t fun.
• Build with resilience and recovery in mind, use in-metro disaster recovery (available on AWS and Azure) and Auto Scaler for Cloud Foundry apps.
• Remember the shared responsibility model, SAP ensures resilience for BTP services, but your solution’s stability is still your job.

I liked how grounded this session felt. No buzzwords, just real-world advice for building scalable, maintainable SAP solutions.

• Why Enterprise Architecture Still Matters by Dominik Rose

Dominik introduced Agentic AI within SAP LeanIX. He talked about how Agentic AI will reshape enterprise architecture as we know it.

• You can automate EA activities. AI Agents execute tasks directly in SAP LeanIX, helping EA become more efficient. 
• Analyse large-scale data. Agents process huge amounts of data to guide the architect through optimisation opportunities.
• Source data automatically from external code repositories to build inventory autonomously. 
• Connect SAP Lean IX data and functions with other business systems for cross-application workflows.

It’s fascinating to see enterprise architecture entering an era of intelligent automation.

Dominik’s message was clear: AI won’t replace architects, but it will amplify them. 

• SAP Business Data Cloud. The Five Guiding Principles

Another great deep dive focused on SAP Business Data Cloud (BDC) and its evolving product strategy. The talk introduced five guiding principles that really define where SAP’s data ecosystem is headed:

1. Persona Centricity – one consistent definition of data, accessible to both IT and business users (and even AI agents).
2. Modern Lakehouse Architecture – scalable, cost-efficient, and designed for both structured and unstructured data.
3. Data Products Economy – pre-built, SAP-managed, metadata-rich data assets ready to consume.
4. Best-in-Class Analytics and AI/ML – native integration with SAC, Datasphere, and AI Foundation.
5. Open Data Ecosystem – true zero-copy, bi-directional data sharing.

For me, this was one of those “aha” sessions; it really clicked how SAP is positioning BDC as the connective tissue for enterprise data.

• Most AI Pilots Fail to Deliver ROI by Liya Spiegel

And then there was Liya Spiegel’s session: short, sharp, and brutally honest. 

Her main point was: don’t do AI because everyone else is doing it.

She walked through several case studies where AI projects failed simply because teams started with a technology rather than a real business problem. 

The takeaway?

“Focus on the problem first, and let the technology follow.”

That line stuck with me. It’s a simple mindset shift, but one that separates successful AI initiatives from wasted effort.

Day 2 & 3: Community Moments & Hands-On Learning

Day 2 kicked off with an insightful interview with Daniel Yu, CMO of SAP BDC, on how developers and consultants can prepare for the BDC revolution. Dan shared some good information regarding the BDC revolution and SAP integration with non-SAP solution partners. I’ll post the interview article with Dan soon. 

Later, I joined a hands-on workshop to build SAP SBPA flows using AI, a few hiccups at first (typical demo gremlins), but it worked beautifully in the end.

As part of AvoTechs, we hosted two sessions on UX Transformation and Clean Core, and both were packed to the door. The energy in the room was fantastic.

Here’s a glimpse of the sessions:

UX Session by me

Clean Core Session by Szymon (our technical SAP Consultant)

The Bigger Picture Behind the Buzz

The pace of innovation at SAP remains impressive. However, the reality is that many organisations still struggle to translate these innovations into their landscapes. Aligning them with modern enterprise architecture principles is not a simple lift-and-shift. One of the biggest challenges I see with my clients is the upskilling gap; architects and developers need new capabilities faster than most organisations can train them.

It’s no longer enough to just issue new guidelines, KPIs, or quality gates. What’s needed are targeted capability-building plans, platform-specific training paths, and, most importantly, a consistent Clean Core commitment that’s enforced through both architecture and culture.

Overall, the direction is undeniably promising. The challenge ahead is for consultants and developers to make all this innovation work in real, complex system landscapes.

That’s my wrap-up from TechEd 2025. If you were there, I’d love to hear what stood out for you, and if you missed it, don’t worry, SAP has already hinted that TechEd will be back next year in Berlin.

This article is originally published at: https://www.linkedin.com/pulse/sap-teched-2025-kebab-capital-world-mateusz-skrzyniarz-hlmef/