Artificial Intelligence
The Action AI Reader
What we’ve published (to date), what we’ve learned so far, and what continues to matter.
Saying that AI is moving quickly is an understatement. The hard part is not keeping up with the product releases, it’s building the conditions under which AI can be trusted, adopted, and sustained.
Here on the Action blog, we’ve been writing about AI the way we work with our clients: highly engaged and curious, systemic and pragmatic, sober and a tad skeptical. This roundup collects our AI-related posts so you can scan the landscape and follow the threads that most interest you.
The AI Reality Check
Signal in the Noise: The 2025 AI Retrospective
If 2024 felt like AI anxiety, 2025 has looked more like operational reality. This retrospective from Shaun Davis frames what changed, what didn’t, and why “doing AI” tends to surface foundational problems in data, governance, and decision-making.
AI vs. Automation: A Useful Line in the Sand
AI vs. Automation
A lot of teams call something “AI” when what they actually need is better automation, stronger workflow design, or clearer definitions. This post clarifies the distinction and offers a practical way to decide which tool fits the job.
Ontology and Grounding: When AI Needs a Model of Your World
Ontology, Model Grounding, and Why It’s Showing Up Everywhere
As AI systems become more useful inside real organizations, they increasingly depend on structure: shared definitions, explicit entities, clean semantics, and a model of how the business actually works. Here, Action’s founder, Keith Helfrich, explores why ontology may be less academic than it sounds.
The Emerging Stack: Data Cloud, Tableau Next, and Agentforce
Salesforce Data Cloud, Tableau Next and Agentforce: A Technical Explainer
The platform layer is shifting. This explainer package of video, article, and downloadable PDF covers what’s new, what’s marketing, and what may be structurally important for analytics teams trying to build AI-enabled workflows that don’t collapse under real-world complexity.
Hiring in the Age of AI
Building Your Team in the Age of AI
As AI lowers many of the technical barriers in analytics and engineering, the advantage shifts from raw skill to judgment. Here, Shaun Davis explores why modern data teams should prioritize discernment, taste and friction reduction when hiring. The core idea: when building solutions becomes easier, the real differentiator is choosing the right problems and finishing the work with clarity and care.
Metadata as the Control Plane
Metadata Moves to the Forefront of Enterprise AI
As enterprise AI shifts from demos to production, metadata is moving from back-office governance to front-line strategy. Keith argues that the real battleground is no longer pipelines and integration, it’s owning the context that makes AI useful: lineage, access control, semantics, and trust. The key warning: AI agents without metadata are just very expensive interns.
Human Flourishing and the AI Moment
My Takeaways from the 2025 Missional AI Conference
At the 2025 Missional AI Conference, the conversation went beyond model releases and technical breakthroughs to a deeper question: how should AI shape human flourishing? Robert Rouse reflects on insights from leaders across industry and research, highlighting themes of trust, ethics, reasoning systems, and the human values embedded in the technologies we build. The through-line: AI is not just a technical challenge, it is a moral and societal one.
The AI Organization Advantage
The AI Cheat Code
In a world obsessed with prompts and productivity hacks, Shaun argues that the real advantage with AI is much simpler: organization. Clear task definitions, explicit “done” criteria and structured workflows help AI produce far better results. The takeaway is refreshingly practical: the same systems that help teams work well together, like user stories and Kanban boards, also help humans and AI collaborate effectively.
Want Help Applying Any of This to Your Situation?
Are you in the middle of an AI initiative and dealing with pain points around data readiness, governance, trust, or adoption? Action can help you diagnose what’s structural vs. what’s tooling.
