AI Over 40 Series - Week 10: Creating Simple Automations with AI

How AI literacy turned a routing nightmare into a three-hour solution

When a key team member went on extended leave, I inherited their workload—including routing every signed client agreement to the right stakeholders. My predecessor’s email-template system looked simple: copy, paste, fill blanks, send. In practice, I lacked the data those templates required, forcing me to hunt through systems, old emails, and shared drives. Agreements that didn’t fit the templates left me guessing about the recipients. The manual workaround sucked up half my morning and frustrated everyone. I jumped into Microsoft Copilot Studio, convinced I’d whip up an AI agent in hours. Instead, I soon drowned in generative-AI code snippets and developer-level configuration. My agent dream stalled, and routing complaints piled up.

The AI agent trap

Too often, we leap straight to “build an AI agent” when we hit a process snag. Platforms tout “citizen developers” and “democratized IT,” but the reality is you end up writing and debugging code—exactly what you hoped to avoid. Agents can help in certain scenarios, but not when you haven’t defined your real problem.

The breakthrough: asking better questions

Finally, I paused and asked: What’s the actual challenge here? It wasn’t automation per se—it was that salespeople held the required information, yet the task landed on my desk. The real fix: shift responsibility, not just technology.

AI literacy in action

With a clearer problem statement, the solution emerged:

1. Intake form for salespeople. Collect agreement details upfront.
2. Dynamic branching. Tailor questions based on agreement type.
3. Automated routing. Fire off emails to appropriate recipients.

I used Microsoft Forms for the dynamic questionnaire and Power Automate for routing and email generation. But I had never built either. Rather than trying to code an AI agent, I turned to Gemini as my “consultant”:

• Form design: Guidance on setting up conditional logic in Forms.
• Flow creation: Step-by-step instructions for building Power Automate workflows.
• Distribution logic: How to map form responses to distribution lists.
• Email automation: Attaching signed agreements and customizing message bodies.
• Troubleshooting: Debugging tips when flows failed.

In three hours, I had a fully functional, transparent, and maintainable solution—no custom code, no tickets to IT, and no AI agent. Most importantly, I learned how to build similar automations myself.

The literacy advantage

This experience highlights why AI literacy is more powerful than AI agency:

• Problem redefinition: AI failures helped me see that process ownership, not technology, was the bottleneck.
• Solution architecture: Once I defined the right problem, AI pinpointed the tools and design patterns to solve it.
• Implementation guidance: AI bridged my knowledge gap on Forms and Automate, turning me from novice to builder in hours.
• Troubleshooting partner: AI assisted in debugging and refining workflows without needing a developer.

The unexpected benefits

Compared to an opaque AI agent, this approach delivers:

• Transparency: The sales team sees and understands the intake form.
• Maintainability: Flows are visible in Power Automate, so anyone can update them.
• Adaptability: Changing agreement types or email templates takes minutes.
• Ownership: Sales owns the data collection; I own the routing process.

And I now have the skills to tackle dozens more automations—agents solve one problem, but literacy multiplies your leverage.

Why this matters now

Too many leaders assume that “AI agency” means AI can read our minds, deduce context, and implement solutions end-to-end. Truth is, until AI masters empathy and bureaucratic nuance, we remain responsible for problem definition, process design, and change management.

No AI agent would have realized that salespeople should fill in key fields. Only by understanding the real issue did I avoid wasted effort and deliver a robust solution.

The world of possibilities

With AI literacy, any leader can:

• Build workflow automations without raising IT tickets
• Create dynamic tools without writing code
• Solve hidden process pain points in hours, not weeks
• Launch pilots that once died in committee

We don’t need AI to do our work; we need AI to teach us how to do new work faster and better.

Your Week 10 Challenge: Find your agreement-routing problem

1. Identify a frustrating process. Something you’ve wished an AI agent could fix.
2. Step back and clarify. Is it a technology gap, a process issue, or an ownership dilemma?
3. Question assumptions. Who holds the necessary data? Who should drive the process?
4. Explore existing tools. Could low-code/no-code platforms handle this if guided correctly?
5. Use AI as your teacher. Pick Gemini, Claude, or ChatGPT and ask how to build the solution—don’t ask it to be the solution.
6. Document your journey. Track time from discovery to deployment and share your lessons.

The bottom line

I spent weeks chasing an AI agent for a problem solved with three hours of AI-guided low-code work. That’s the power of literacy before agency: AI doesn’t just do tasks—it teaches us to do tasks ourselves.

This post is part of our “AI Over 40” series. It first appeared on LinkedIn: AI for the Over 40 – Week 10: Creating Simple Automations with AI.

Next Week: We’ll examine another case of how AI literacy transforms solution design—and the exponential impact it can have on business performance.

Read more AI and Copilot blogs.

Trending Posts