Did Generative AI Kill RPA?
With the rapid advancements in Generative AI (GenAI), an interesting debate is emerging in the field of business automation: Will GenAI render traditional Robotic Process Automation (RPA) bots obsolete, much like the car replaced the horse and cart? Or will it lead to a resurgence by enhancing automation with systems that can act independently and make decisions without human intervention—often referred to as agentic automation?
Potential Disruption of RPA by GenAI
Traditional RPA excels at automating mature, deterministic processes involving structured data and a limited number of systems. However, RPA bots are often brittle—they can 'break' when processes or systems change, requiring reprogramming. This is problematic, especially when organizations expect bots to handle repetitive tasks without human intervention and struggle when the bots are offline.
GenAI introduces the ability to handle non-deterministic processes and unstructured data, plus brings in other multi-modal capabilities—text, voice, pictures, etc. AI introduces confidence levels; if the model is very confident it has made the right decision, it can continue and complete without human touch. Less confident decisions may need to be referred to human teams to process. Either way, a strong governance and quality assurance (QA) capability is required.
How do you see GenAI impacting your organization's automation strategies?
Combining GenAI with RPA Leading to a Resurgence
Alternatively, combining GenAI with RPA could lead to a renaissance in deterministic automation. This integration enhances RPA's capabilities, making bots more adaptable and less prone to breaking due to process changes. By incorporating GenAI, bots can better interpret data, adjust to variations, and execute higher order tasks more reliably.
At Telstra in 2018 (pre-GenAI era), we hypothesized that machine learning and process diagnostics could be used to detect process changes impacting bots, allowing developers to diagnose and reprogram bots quickly, reducing business impacts if they were offline. Today, RPA software + GenAI can diagnose process changes and automatically reprogram bots in certain circumstances. Some providers offer self-healing bots that adapt in real-time, minimizing downtime and maintenance efforts. For instance, several companies have implemented self-healing bots powered by GenAI, significantly reducing their bot downtime.
领英推荐
Opportunities
Challenges
In conclusion, while the rise of agentic automation powered by GenAI is evident in the short to medium term, its full potential remains to be seen. Businesses must weigh the benefits against the challenges to decide whether to lead the charge or strategically follow. What path will your organization take?
#GenerativeAI #RPA #Automation #AIInnovation #BusinessStrategy
?
very interesting announcement from Anthropic 'Available?today on the API, developers can direct Claude to use computers the way people do—by looking at a screen, moving a cursor, clicking buttons, and typing text.". Take a look https://www.anthropic.com/news/3-5-models-and-computer-use?campaignId=11442233&source=i_email&medium=email&content=Oct2024Sonnet&messageTypeId=140367&_sm_au_=iVVMTtM0jqND477F
Helping community, colleagues and clients to "Grow for tomorrow"
5 个月Nice insights Tim Barnes. I agree RPA has been obsolete now from a long time. Since last 18 months back -we have been integrating RPA (both UI and API based automation) with GenAI. You can access the case studies here at https://www.mindfieldsglobal.com/generative-ai
CEO & Founder @ Roxus | Digital Transformation | Intelligent Automation
5 个月Nice share Tim Barnes. The debate between GenAI and RPA isn't a competition in my view, but rather a convergence. By integrating GenAI with RPA, enterprises can address the limitations of both technologies—especially when it comes to agility and handling semi/unstructured data and processes. GenAI undoubtedly opens exciting new horizons, but one key challenge is managing its ‘hallucinations’—especially in environments that demand accuracy. At Roxus we’re offering clients a unique approach with Attended Assistant's to deliver “Hybrid Automation”; curating interaction with GenAI and AI solutions. This helps create a more resilient automation system, while maintaining a strong foundation of governance and human oversight. Would love to show you what we're doing some time.
Data Scientist/Generative AI @ ANZ | LinkedIn Top Data Science Voice | Advanced Analytics | Generative AI | Machine Learning | Journal Reviewer with Elsevier | Innovator | Risk Tech Enthusiast
5 个月I agree with you, as businesses navigate this transformative landscape, they must carefully consider how to leverage the strengths of both GenAI and RPA to drive innovation and maintain a competitive edge in a rapidly changing market. The future of automation lies in a synergistic approach that combines the best of both worlds, enabling organizations to thrive in an increasingly digital environment and thanks to our OPEN Source Community which is helping organisations to evolve quickly rather than relying on tech Giant's.
Advanced Analytics, AI, ML and BPM Researchers and Advisors | Scaling Advanced analytics COEs | Digital GCCs | ex PWC M&A |ex CBA GSD |ex E-Lance (now Upwork)|ex ICICI | IIT-IIM | AICD
5 个月Tim Barnes RPA was a "band aid" measure even during its heydays when every Big Four firm was touting it as if it were manna from the heavens! ML was always going to make it redundant and now with the advent of GenAI bots, those old clunky NLP bots that RPA relied on (and still does in most instances) are as outdated as those old black and white TV sets with circular dials ?? You only need to look at UiPath's stock price to know the fate of RPA