Swag Discussion Series | Week 1

What's one assumption about automation you're rethinking after this week?

I always assumed that automation was only about automating the regular, mundane and repetitive tasks and using AI along with RPA would just help in increasing the efficiency of the existing bots, but after this week, this assumption no longer looks valid to me as we are now capable of not just optimizing the existing bots but we are now capable of automating those processes as well which were NOT touched before due to decision-making and complex scenarios, thanks to AI capabilities!

I was under the impression that automation was about automating repetitive tasks and the hours saved from the automation. It’s more than that...it’s understanding the why...and determining the best to address it. 

One assumption I'm rethinking is that automation starts with building bots. This week helped me realize that successful agentic automation starts with understanding the business problem, defining outcomes, and designing the right workflow before choosing any technology. The Agentic Design Framework showed me that good design is more important than jumping straight into development.

One assumption I'm rethinking is that automation starts with the process itself. This week reinforced that successful agentic automation starts with understanding the business problem and desired outcome first. The Agentic Design Framework helped me see the importance of breaking down the objective, identifying where human judgment is needed, and determining which tasks are best handled by AI agents versus traditional automation. Instead of asking, "What steps can I automate?" I'm starting to ask, "What outcome am I trying to achieve, and what's the best combination of people, AI, and automation to get there?" I think this mindset will lead to solutions that are more scalable, resilient, and valuable to the business.

One assumption I'm rethinking after Week 1 is that automation's primary purpose is to save time by replicating human actions.

In my projects, I've often looked at a repetitive task and immediately started thinking about how to automate it. My mindset was, "If a person clicks these buttons every day, a bot should do it instead." But the Agentic Design Framework made me realize that this approach can sometimes automate inefficiencies rather than solve the actual problem.

A real-life example is expense claim processing. At first glance, it seems like a perfect candidate for automation because employees submit forms, managers approve them, and finance processes payments. Earlier, I would have focused on automating each step faster. Now, I would first ask questions like: Why are claims delayed? What information is usually missing? Which approvals create bottlenecks? Are all claims really following the same path?

By breaking down the process before building, we might discover that the biggest issue isn't manual effort but inconsistent data, unnecessary approvals, or lack of decision-making support. In that case, an agentic solution could intelligently validate information, route exceptions, and assist decisions rather than simply mimicking clicks.

My biggest takeaway from Week 1 is that great automation doesn't start with a bot, it starts with understanding the business problem. The best solutions come from clarity, process deconstruction, and thoughtful design. Building is important, but building the right solution is what creates real value.

One assumption I'm rethinking is that automation should primarily mirror the existing process. Exploring agentic process automation has shown me that the real value comes from reimagining work around outcomes rather than tasks. By deconstructing processes from an agentic perspective, we can identify opportunities to simplify decisions, eliminate unnecessary handoffs, and optimize workflows before automation is even designed. The result isn't just a faster process, but a smarter one.

I used to think that automation was just recording repetitive human clicking actions.

After the Week 1 courses, Now I realize that true automation is setting goals for AI and letting it find its own solutions. That’s amazing!

Week 1 challenged a mindset I've carried throughout many automation projects: "If a process is manual, automate it."

The Agentic Design Framework demonstrated that automation should begin with curiosity, not technology. Instead of immediately focusing on tasks and workflows, it encourages us to understand objectives, constraints, decision points, and the real business problem.

For example, in procurement workflows, delays are often assumed to be caused by manual approvals. My earlier approach would have been to automate approval routing. Now, I would first investigate whether delays stem from missing information, unclear ownership, policy exceptions, or inconsistent decision-making.

This perspective shifts automation from task execution to intelligent problem-solving. Rather than replicating human clicks, agentic solutions can analyze context, guide decisions, and adapt to changing scenarios.

My biggest insight from Week 1 is that successful automation is not measured by how many steps are automated, but by how effectively the solution improves business outcomes. The most valuable automation begins with understanding, not implementation.

So, ​@Lu.Hunnicutt ,

I used to think automation started with a workflow. Week 1 changed that, it starts with the business problem. APA combines orchestration, reasoning and execution to deliver outcomes, not just automation tasks. As it is stated, AI isn’t changing how we build solutions, it is changing who can build them. That mindset shifts really resonated with me….