#Mindcandy are scenario prompts to help companies formulate strategy. Find out more here.
- If you need help with developing these prompts, or want to shoot the breeze and talk about books, strategy, leadership, innovation and future trends, you can book a free appointment here.
- If you want me to speak at your event, or book me as a coach, contact me at ron@ronimmink.com
- If you want to receive all near daily mind candies, you can subscribe here https://scenarioprompts.substack.com/.
Digital transformation #mindcandy 29 May 2025
Today’s topics are RAG, data architecture, predictive analytics, ERP, IoT, IBM automation, AI agents, AI insubordination, and the prompts frontier.
RAG best practices for enterprise AI teams
https://www.techtarget.com/searchenterpriseai/tip/RAG-best-practices-for-enterprise-AI-teams
Many organizations are already using commercial chatbots driven by large language models, but adding retrieval-augmented generation (RAG) can improve accuracy and personalization.
Data Vault makes risk and finance at Rabobank more agile and scalable
https://www.deloitte.com/nl/en/what-we-do/case-studies/rabobank.html
Data architecture deserves everyone’s undivided attention
The Democratization Of Predictive Analytics
Advances in AI and machine learning are transforming predictive analytics from an elite capability into a widely accessible tool. What was once siloed in data science departments is now being placed directly into the hands of business teams.
Is AI the Death of ERP? Not Quite. But It Might Just Kill the Front End.
https://erp.today/is-ai-the-death-of-erp-not-quite-but-it-might-just-kill-the-front-end/
As enterprises adopt AI tools, there’s a critical need to balance innovation with robust governance, ensuring that security, data integrity, and compliance risks are addressed even as they explore agent-driven workflows.
IoT and connectivity choices: not just about cost
https://evertiq.com/news/2025-05-27-iot-and-connectivity-choices-not-just-about-cost
The Internet of Things (IoT) is no longer the domain of futuristic visions — it has become a tangible, widespread element of everyday life, from smart homes and modern cities to industrial and logistics applications.
From copilots to orchestrators: AI agents reshape IBM’s automation playbook
IBM Corp. is focused on developing AI agents that execute across systems rather than merely assist at the edges. These agents are designed to integrate with legacy and modern tools, orchestrating processes across the full sprawl of enterprise infrastructure.
Can AI Agents Be Trusted?
https://hbr.org/2025/05/can-ai-agents-be-trusted
As agents gain decision-making power, they also introduce significant trust and accountability risks, including susceptibility to hacking, marketing manipulation, misinformation, and conflicting loyalties
Advanced OpenAI Model Caught Sabotaging Code Intended to Shut It Down
https://futurism.com/openai-model-sabotage-shutdown-code
We are reaching alarming levels of AI insubordination.
The prompt frontier—how engineers are learning to speak AI
https://www.engineering.com/the-prompt-frontier-how-engineers-are-learning-to-speak-ai/
AI models require well-structured prompts to perform optimally. A poorly phrased question or vague instructions can lead to suboptimal or misleading outputs.
