Knowledge that matters
Stay informed on our blog and news platform.
Optimistic Real Pessimism · A manifesto for AI engineering
Seven articles for everyone who wants to build AI automation for real — beyond demo hype and omnipotence fantasies. Five pillars, one manifesto, grounded in practice.
Part 1 of 7
The Manifest – A Wake-Up Call for AI Automation
June 10, 2026
Still believe in AI miracles? Then you've never watched an agent stuck in an infinite loop. Our manifesto for everyone who actually builds — welcome to optimistic real pessimism.
Read morePart 2 of 7
Probability Not Commands – Guardrails for Probabilistic Code
June 17, 2026
Software used to be predictable: if X, then Y. AI is different: if X, then probably Y. Treating AI like legacy RPA demands guardrails — here's how we build them.
Read more
Part 3 of 7
When the LLM Drowns in Data Chaos – RAG, GraphRAG and Context Engineering
June 7, 2026
Blog 02 demonstrated how an AI agent becomes reliable through clear data structures and validation loops: it understands in a structured way what is required. But even a correctly configured agent will return incorrect…
Read more
Part 4 of 7
When Agents Run Wild – State, Memory and Cost Kill-Switches
June 8, 2026
Blog 03 demonstrated how an AI agent finds the right information: Through targeted pre-selection and a network of relationships between business objects, the model receives exactly what applies to a specific customer…
Read more
Part 5 of 7
Privacy by Design – Hybrid Architectures for Enterprise AI
June 15, 2026
The most powerful language models run in the cloud. The most sensitive corporate data — customer names, contract terms, internal business information — must not easily reach external services. This seems like a…
Read more
Part 6 of 7
Why Your AI Might Be Worse Tomorrow – LLMOps and Model Drift
June 22, 2026
Classic program code behaves predictably: if you don't change the source code, you get the same result. This is not true for AI systems. Language models are regularly updated by providers in the background. The same…
Read more
Part 7 of 7
Permissions and Governance – Who Owns the Agent?
June 29, 2026
The previous blogs described the technical architecture: structured outputs, targeted information selection, execution control, data protection, and quality assurance. What's still missing is the organizational side:…
Read moreSAP Business AI Platform · series
Seven related articles on roles, architecture, and diagrams around the SAP Business AI Platform.
Part 1 of 7
The SAP Business AI Platform Is Here: But Who Will Actually Operate It? 🚀
May 19, 2026
Introduction to a blog series on the SAP Business AI Platform: four new key roles, legal guardrails, and and why the real challenge is the organisational transformation.
Read morePart 2 of 7
Deep Dive Part 1: The Intent Architect – The Lawmaker of Autonomous AI Agents
May 21, 2026
First deep dive into the SAP Business AI Platform: what an Intent Architect actually does in Joule Studio — from deterministic tool provisioning to legally enforceable guardrails.
Read morePart 3 of 7
Deep Dive Part 2: The AI Agent Supervisor / Exception Manager – From Clerk to Air Traffic Controller
May 23, 2026
Second deep dive into the SAP Business AI Platform: how the role shifts from data-entry clerk to exception manager — monitoring, KPIs, and legally sound approvals.
Read morePart 4 of 7
Deep Dive Part 3: The Enterprise Semantic Engineer – Guardian of the Digital Memory
May 26, 2026
Third deep dive into the SAP Business AI Platform: Knowledge Graph, BDC, and ontologies — why semantics trump raw data and how the context architect keeps it safe.
Read morePart 5 of 7
Deep Dive Part 4: The AI Governance & Compliance Officer – Architect of Legally Sound AI Guardrails
May 29, 2026
Fourth deep dive into the SAP Business AI Platform: EU AI Act, IAM for agents, API policies, and Explainable AI — why governance is the enabler of autonomous systems.
Read morePart 6 of 7
Deep Dive Part 5: Administrators & Developers – The Foundation of the Autonomous SAP Enterprise
June 2, 2026
Fifth part of the series: how BTP, integration, ABAP, and CAP roles grow alongside AI agents — infrastructure, Clean Core, and agentic engineering.
Read more
Part 7 of 7
Deep Dive Part 6: Architecture in Diagrams – Layer Model and Process Flow
June 5, 2026
Two diagrams illustrating the SAP Business AI Platform: from the engagement to infrastructure layer, and the timeline from design-time through run-time to exception handling.
Read moreMore articles
Where Theory Meets Reality: The Next Generation of SAP Experts in Chișinău
April 8, 2026
A look back at three intensive days of solution architecture, data strategy, and hands-on innovation at UTM.
Read more
A sovereign learning platform: Why we use Nextcloud and PortKnox
10 April 2024
Tired of the data jungle? If you want a learning platform as intuitive as Microsoft Teams but fully GDPR-compliant, here is how Nextcloud, automated API scripts and hosting with PortKnox give you one central environment — already proven from Moldova to Armenia.
Read more