From RPA to AI Agents: A Journey Through the Layers of Intelligent Automation

By Subhabrata Pal

There has never been a bigger force for change than technology. It shifts the way we live, interact, and do business—often before we even realize what’s changed. At our latest knowledge-sharing session, we took a deep dive into this very evolution—tracing the path from Robotic Process Automation (RPA) to Intelligent Automation (IA) and now to AI Agents.

It wasn’t just another theory-heavy discussion. We watched compelling videos, exchanged stories, and grounded every concept in relatable, real-world examples. This format—part workshop, part fireside—helped decode the invisible transformation happening around us.

RPA: The Foundation Layer

RPA thrives in rule-based environments. Think: repetitive, structured tasks like data entry, payroll processing, or invoice reconciliation. These are jobs that don't require human intuition—just strict adherence to pre-defined rules.

➡️ Example: Years ago, training attendance was logged manually on paper. Today, a scan of an employee ID feeds directly into Google Forms and enterprise systems. That's RPA in action—efficient, reliable, and repeatable.

Intelligent Automation: Adding Cognitive Muscle

But what happens when data is unstructured? What if the input isn’t a spreadsheet but a scanned prescription, a customer complaint email, or facial expression analysis?

This is where Intelligent Automation steps in—an evolved form of RPA that marries basic automation with AI and ML capabilities. It can process vision, voice, and language to make basic decisions, adapt over time, and even handle exceptions.

➡️ Example: Static code analyzers can now flag issues in real-time. NLP-driven bots can draft blogs from transcripts (as I recently did after moderating a panel!). The shift from humans to “smart bots” is accelerating.

AI Agents: Towards Autonomy

AI Agents represent the convergence of automation, intelligence, adaptability, and autonomy. They not only perform tasks but also learn, reason, and predict.

💡 Where Do We Go From Here?

As we explored in the session, not every problem needs an AI Agent. RPA still has its place, especially for low-cost, repetitive tasks. IA works well when semi-structured data needs intelligent handling. AI Agents, while powerful, are best suited for dynamic environments and higher-level decision-making.

“AI won’t replace you, but a person using AI might.”

Let’s focus less on job displacement and more on skill transformation. After all, AI may automate tasks, but it amplifies human potential.

🌐 Looking Ahead

We’re excited about what’s coming next: fireside chats with leaders, panel discussions, and practical showcases. As we diversify formats, one thing remains constant—our focus on learning, evolving, and staying ahead of the curve.

Some Anecdotes shared during the presentation.

🔹 1. Manual Attendance Logging vs. Bulk Excel Uploads

Just a few years ago, classroom attendance was logged manually using sheets and tables. This data was then transferred to digital systems by data entry clerks—a repetitive, mundane task.

Now, attendance is handled seamlessly through Excel uploads, ID scanning, and automated form generation. This is considered a classic use case of RPA.

🔹 2. Blog Creation from Meeting Transcript Using ChatGPT

After a panel discussion was moderated, real-time note-taking was found to be a challenge. Later, a Zoom transcript was used along with ChatGPT to generate a complete blog, including summaries and action items. This anecdote is seen as an example of intelligent automation and AI-assisted content generation.

🔹 3. iPhone Autogenerating Trip Videos

During an international trip, videos of the visit were automatically compiled by the iPhone, curated from the best daily photos. Hours of manual video creation were saved, making this a strong example of how AI agents are used to deliver value proactively through embedded intelligence.

🔹 4. Initial Panic Around RPA and Job Loss

When RPA and BPA were introduced 8–10 years ago, fears about job loss were widely felt. However, over time, those same professionals were seen adapting by learning how to code or manage bots—highlighting how evolving roles are shaped by technological change.

🔹 5. Immigration at JFK Airport – Manual to Self-Service Kiosks

During an earlier visit to the U.S., immigration was handled manually. Just six months later, most processes had been shifted to automated self-service kiosks with minimal human involvement—demonstrating how quickly customer-facing operations can be reshaped by automation.

🔹 6. Tesla Fixing a Software Defect Remotely

As shared by your company president, a suspension issue in a friend’s Tesla was experienced while driving over road bumps. The problem was expected to be fixed at a workshop. Instead, it was diagnosed remotely by Tesla and was resolved via a software update—without the car being brought in. This scenario is viewed as a powerful example of how AI Agents, IoT, and OTA updates can be used to redefine service experiences.

Tesla isn’t just a car company. It’s a software company on wheels, showcasing what’s possible when AI is deeply embedded into everyday functions.

🔹 7. IoT Surveillance by Friends Abroad

During a visit, friends from Dubai demonstrated how their children in India were monitored through CCTV feeds connected via Wi-Fi. This setup was enabled by instrumented, interconnected, and intelligent systems—the foundation of IoT-driven intelligent automation.

🔹 8. Static Code Analyzer in Software Projects

It was shared that tools like Gherkin and Cucumber now are used to automate the creation of acceptance tests, replacing manually written test cases. Syntax and structural issues are flagged by static code analyzers, improving code quality and reducing manual effort—once again tying into the value delivered by RPA and IA.

If you found these insights valuable, share this post, drop a comment, or let us know how you’re navigating the RPA → IA → AI Agent journey in your domain.

Thanks Dola Pal swati banerjee for sharing your thoughts.

Youtube Link - https://youtu.be/NK28fD7qEKQ