Swag Discussion Series | Week 2

I used to think smarter automation meant adding more logic. Now I think it means giving the right capability to the right task. Turns out, the best workflow for any process isn't the one with the most automation - it's the one where every component knows exactly what it's responsible for.

Still in the Foundational Automations section, but I already came across the Structured Data Extraction feature in the Recorder command which is completely new to me, and I found it really cool.

Earlier, I had to find the number of rows/ items, build dynamic DomXPaths, loop and extract the data for each item, add it to a table, and handle scrolling if needed. Now, all of that can be done in a single step.

It's a real time-saver and makes it much simpler.

As an RPA developer, this week reinforced that the future isn't about building larger automations it's about orchestrating specialized capabilities. Instead of forcing every scenario into workflow logic, we can leverage AI Skills, Document Automation, APIs, and RAG where they add the most value. That shift changes not only how I design automations, but also how I identify automation opportunities.

Invoice processing looks completely different to me now. Before, I thought it was mainly rule-based RPA. After learning Task Automation, APIs, Document Automation, AI Skills, and RAG, I now see how AI can understand documents, use business knowledge, and make smarter decision not just automate repetitive task.

I looked at processes by asking, "What repetitive steps can I automate?" Now I ask, "What decisions require context, and how can AI provide that context?"

Features like Document Automation, AI Skills, API Tasks, and RAG showed me that modern automation isn't just about executing workflows. It's about understanding information, retrieving relevant knowledge, and making informed decisions before taking action.

This has completely changed how I evaluate automation opportunities. The future isn't just faster bots—it's intelligent agents that collaborate with people to solve business problems.

My mind is in overdrive now :) 

I like the concept of one agent for 1 task and then having an orchestrating agent.  The art of the possibilities!  Looking forward to possibility uses agents for administrative tasks!

One process that looks very different to me now is a workflow we use to investigate insurance coverage for GLP-1 medications before a patient visit. Today, staff spend time checking eligibility, running test claims, reviewing coverage details, and documenting the results after a patient's appointment, which can delay decisions and follow-up actions. 

A few weeks ago, I would have looked at this as a series of manual tasks that could be automated. Now I see it as a business outcome we're trying to achieve: giving staff a complete coverage picture before the patient's appointment. That shift has me thinking less about individual tasks and more about how data, AI, automation, and people can work together to provide the right information at the right time, with humans focused on exceptions and judgment calls.

After Week 2, I see many everyday business processes differently. What once looked like a series of manual steps now feels like a connected workflow that can intelligently handle data, communicate through APIs, understand documents, and even retrieve the right information using RAG. Instead of asking, "Who has to do this task?" I now find myself asking, "How can AI and automation work together to do this more efficiently?" It has completely changed the way I look at process improvement.