Enter the new era of automation with AI in RPA – In conversation with UiPath MVPs

UiPath’s Most Valuable Professional (MVP) Award is the highest recognition to its community members for their outstanding contribution, innovation, and evangelism shown in the larger automation community.
Enter the new era of automation with AI in RPA - In conversation with UiPath MVPs

As per a report by Gartner, through 2024, large organisations will triple the capacity of their existing Robotic Process Automation (RPA) portfolios. It also predicts that 90 per cent of large organisations globally will have adopted RPA in some form by 2022, as they will look to digitally empower critical business processes through resilience and scalability while recalibrating human labour and manual effort.

UiPath is one of the biggest names in the field of RPA. UiPath’s Most Valuable Professional (MVP) Award is the highest recognition to its community members for their outstanding contribution, innovation, and evangelism shown in the larger automation community.

In an email interaction, Nisarg Kadam, UiPath MVP 2021, working as a Senior Consultant at WonderBotz and Sishira Mishra, UiPath MVP 2021, working as Automation Architect at HP, discuss how they started their RPA journey, how UiPath helped them in attaining and honing their skills and what can we expect in the RPA domain in the future.

THE BELAMY

Sign up for your weekly dose of what's up in emerging technology.

RPA journey 

Kadam says that his RPA journey started in 2018 when he learned from the UiPath Academy. Today, he strives to give back the information he learned over the years to the community by helping ensure that everyone knows how to utilise the product and scale the automation beyond what they have imagined. “In order to reach more people, I have started a YouTube channel through which I have shared 131 videos to date, with over 340,000 views and 5,300 subscribers,” he adds.

Over the years, Kadam has participated in and won two global hackathons where he showcased the power of ML integrated RPA using UiPath’s AI Center. He has also been a speaker in over 20 UiPath Community sessions, most recently with his project – ‘Automation Pathshala’ where he trained how to use the UiPath application to over 1300 registered participants.

Mishra’s RPA journey started in 2015, working with several vendors when the focus was more on unattended automation with a classic RPA approach. This journey has shifted to cognitive RPA over the years with an increased focus on enabling digital transformation. 

“Over the years, I have been involved in multiple automation development projects with varying complexity across business functions. Having come from an IT background, I did not understand business processes in-depth. However, working with RPA, I have been able to realise business problems and how this technology can be a key solution while maximising innovation.” –  Sishira Mishra 

He has had the opportunity to work on the end-to-end automation ecosystem that involves setting up the infrastructure, discovering potential automation opportunities, designing and developing solutions, and providing support for the production platform. Parallelly, he has worked closely with the UiPath Engineering and product team through the UiPath Insider Program, where he shared ideas and experiences with the UiPath Insider Community.

Scope of automation is increasing beyond imagination

Kadam feels that the RPA industry is growing rapidly, and with this, the scope of automation is increasing beyond our imagination. He has learnt a lot in less than four years but still feels that he knows very little about this industry and has so much more to learn. This shows how huge this industry has become.

In 2018, RPA began mimicking humans. Eventually, when the industry understood the limitations of definite step automation, vendors started exploring cognitive automation by plugging in AI and ML. As a result, they open unlimited doors for RPA to scale in any industry. 

With cloud deployment becoming a trend, RPA will have a significant role in ERP automation, providing processed and less complex data directly from the ERP, thereby improving customer experience. 

This year will be amazing because of. Over the next couple of years, powerful automation discovery products like Task Mining and Process Miningare going to help companies unearth the scope for automation across different domains.

Citizen-led Automation

Mishra says that with ‘Low-code’ and ‘No-code’ applications existing today, everyone can develop robots without a programming background or experience. A key enabler is in this process is Citizen-led Automation. This mindset is taking the industry to the next level, where businesses can produce more value by encouraging automation from operational teams. Products like UiPath StudioX is a true no-code platform that helps business users develop their task automation without the involvement of CoE Developers.

“Low-code and No-code applications are the key enablers in Citizen-led Automation”

This is also helping build a Citizen Developer ecosystem at a large scale across industries. We will soon witness organisations encouraging their Citizen Developers to build task-based automation and improve productivity. With products like UiPath Process Mining and UiPath Task Mining, it is easier for companies to find potential areas of opportunity where CXOs can focus on developing and deploying automation.

Challenges encountered

Mishra remembers the challenge he initially faced to learn automation technologies and receive the right support. With UiPath Academy and UiPath Forum, everything is available for free and at an industry-standard level. Every user can be part of the community, receive access to all UiPath products, learn using video-based training, and receive support from the forum. 

“In our industry, it is important to be up-to-date on all upcoming products. By actively participating in the UiPath Insider Program and connecting over the UiPath Forum, I could collect information about upcoming products and collaborate with the product team to test features over private previews before their release to the general public.”  –  Sishira Mishra

Kadam remembers the most challenging integrating the Python ML model with UiPath when the AI Center had not yet been released. As a result, he had to learn ML with Python and invoke Python prediction models within automation to make it a seamless experience on his projects. 

Kadam also faced operational challenges including setting up a cloud-based document review without human dependency, training bots to read and extract data such as total, tax, and vendor names from 50 different types of invoices, enabling automated email replies, etc. that he was able to resolve using the UiPath AI Center. 

Working with chatbots has been one of Kadam’s most enjoyable experiences. However, he still counts it as a challenge because he learned a completely new product – Google Dialog flow – to integrate a chatbot with RPA. When deployed on the production website, he could not have been happier to see his first chatbot and RPA integration.

Centrally manage ideas and enable a single platform

Mishra says that once we begin scaling automation, one of the challenges we face is centrally managing ideas and enabling a single platform where everyone can share their thoughts and collaborate. This can be easily overcome with the UiPath Automation Hub. 

“The UiPath Automation Hub is a single platform with features to overcome these challenges. Its capabilities not only enable you to submit ideas but also provide several other components including those that help users manage their automation funnel or pipeline easily, create automated documents using Task Capture, consume reusable modules etc.,” he added.

Does not demand too much process revision

RPA as technology is one of the easiest top layer technologies to set up without impacting the company’s current infrastructure. It does not demand too much process revision and adapts rapidly across industries. 

For instance, over the last four years, Kadam has worked with companies in retail, telecom, healthcare, banking, news channels, gaming, education, auction houses, airlines, petroleum, and construction domains to provide successful pilot robots and beautiful end to end automation. For example, when Kadam was working with his auction house client, he realised every single process of an online auction house could be automated easily. However, RPA has not yet evolved to 100 per cent of its capability, and there is a long way to go and grow. 

“Industries continue to focus on RPA adoption to reduce mundane, repetitive work, freeing manual hours and thereby helping users contribute more towards innovation. Banking and financial services, healthcare, and insurance sectors have actively adopted RPA. The manufacturing, retail and telecom industry can also be seen adopting RPA for quality and efficiency improvements in operations to help improve workforce and customer relationships,” added Mishra.

AI in RPA is the next era in the Automation Industry

Kadam says that RPA was introduced in this industry as a technology that automates repetitive tasks, but when it started growing, it was understood that it could also automate cognitive tasks, make decisions and grow smarter with the help of a historical learning feed.

“I was allowed to eliminate tedious manual tasks the L1 support team performed in a service ticket management department. The task was to assign those tickets to an appropriate resolver group based on the availability and priority of the ticket. With the help of the UiPath Action Center, I was able to automate 90 per cent of the process with hybrid robots (90 per cent unattended and 10 per cent attended bots). The fun part of this process was that we were able to predict the category, priority, and impact of the service ticket based on the email body, quite easily with the help of UiPath AI Center, using its ‘English text classification model.” – Nisarg Kadam

Need to go above the rule-based model to a self-decision-making model

Mishra says that AI in RPA is the next era in the automation industry. Robots need to go above the rule-based model to a self-decision-making model to enable more value-added work and benefits. However, this is only possible when artificial intelligence is combined with RPA. With this, in the future, we will be able to move away from traditional RPA to a cognitive RPA. “In today’s ecosystem, with UiPath AI Center, Form AI and Document Understating, we have enabled a business user without any technical background to create their ML model and begin managing unstructured data,” he added.

UiPath AI Center 

Kadam says that when he would hear terms like AI and ML, all he would visualise was a black screen with Python coding on Ubuntu by skilled developers. However, he was mesmerised by the simple drag and drop AI models when he completed his first AI-enabled automation pilot robot within four days of development without learning a single line of Python code. 

“When we delivered our first AI-enabled invoice processing solution, the AI Center helped us gain confidence with over 90% accuracy,” said Kadam. “The feature where an AI Model is trained by learnings from experience and parameters to increase is useful as it helps increase its confidence and become smarter day by day.” 

Mishra adds that The UiPath AI Center is a combination of everything that enables the deployment of an AI-based project. For example, a typical life cycle of an AI Project involves data labelling, training, deploying, and retraining, which can all be performed using the UiPath AI Center. In addition, when teams face typical machine learning challenges, the AI Center helps create an end-to-end AI automation workflow by:

  • Collecting datasets that can be part of data training to enhance the out-of-box model or reuse existing models developed in-house.
  • Performing data labelling within AI Center to create a labelled dataset for ML Training.
  • Running training/ evaluation pipeline by reusing the base model.
  • Deploying the skill via AI Robot, which manages the required resources remotely.
  • Pushing the model to the regular workflow for use on being deployed. 

In some scenarios requiring human intervention in the loop, the Action Center is used to perform human validation and plan the required retraining pipeline to improve accuracy.

Drag and drop an ML model into their workflow

Nisarg adds that the UiPath AI Center allows one to drag and drop an ML model into their workflow without writing a single line of code. It provides:

  • 19 out-of-the-box ML models.
  • over 20 UiPath in-house ML models.
  • and the option to upload the user’s custom-built model.

These models are offered as pre-trained, re-trainable, and non-re-trainable model types. In addition, the UiPath AI Center also offers 16 different in-house document processing models like invoices, receipts, etc., that are also re-trainable to suit the users’ documents and can manage all types of ML models, right from language analysis and image analysis to tabular data processing.

It is easy to use and offers a step-by-step experience that helps users choose an ML model and deploy it without hassle. All the developers have to do is create a project, upload a test/ train data (only in case of re-trainable models), choose the ML model, execute the train and evaluate pipeline (in case of retraining), create ML skill, and drag and drop the same into the UiPath Studio. 

“The AI Center supports the infrastructure to manage training, evaluation, and deployment of models with two special types of robots called the AI Robot and the AI Robot Prov (GPU based),” added Mishra. “While the AI Center manages the complete backend load, the custom or out-of-box models are available to RPA Developers with a simple drag and drop activity. Moreover, it is simple to use with its no-code platform.”

Interesting use cases where UiPath AI Center has helped apply AI and Machine Learning models for automated processes

Kadam has recently written a UiPath Community blog about possible use cases with only ML models and shares some of his hackathons and community projects.

He worked on an AI-enabled customer feedback automation for an Automobile Service centre looking to capture handwritten feedback. The feedback would then have to be classified as ‘Positive’ or ‘Negative’ and uploaded on social media, along with an automated email follow-up to the customer before a service executive reaches out. The project used UiPath Studio, UiPath AI Center, UiPath Automation Hub, UiPath Document Understanding, UiPath Orchestrator, and UiPath Task Capture to help improve customer experience.

He adds, “Automation’s applications are diverse. For example, we can also work on assigning tickets on ‘Service Now’ using ‘English Text classification model’, predict brain stroke using ‘TPOT auto ML classification model’, or build a slack channel chatbot for foreign languages using ‘Language detection’ and ‘language translation’ ML model.”

He has also worked on counting the number of allowed people at a time in a supermarket using the ‘Object detection ML model’, helped maintain old and valuable books by converting handwritten hard copies to soft copies using ‘Document Understanding OCR’ and helped detect entries of trucks and create a gate pass for drivers using the ‘Object detection ML model’. The bot would detect number plates and create auto-generated gate passes for drivers. 

Mishra has used the UiPath Document Understanding models, where the bot reads invoices with varying formats, extracts all relevant information and posts the same into the company’s ERP system. He has also worked with marketing teams with campaigns to validate image-based objects against key deliverables. The AI Centre also becomes useful for SOW validation, email sentiment and intent analysis, text translator capability when working with multiple geographies, GST validation for Indian invoices, etc. 

The UiPath Community

Kadam joined the UiPath Community with the aim of learning and sharing. He shares that The UiPath Community is one of the best as it helps with a lot of support, knowledge, tips, insights, tricks, techniques, codes, ready-made workflows, best practices, hackathon opportunities, job opportunities, career guidance, internship opportunities, use-case repositories, templates, presentations, live sessions, one-on-one with experts, and others. To join, all you have to do is connect with your local chapter on community.uipath.com or join on forum.uipath.com. 

Mishra adds that there are several such initiatives one can be a part of, such as joining the UiPath Academy to learn UiPath Products and technology and the UiPath Forum to communicate and collaborate over issues or concerns and updates on UiPath products. In addition, one can also connect to UiPath MVPs, become an RPA Associate, or receive a UiPath Certified Advanced RPA Developer (UiARD) certification, and find a mentor using the UiPath mentorship program. 

Focus on learning from experience and practical learning

Nisarg and Mishra share some tips one can follow while starting in this industry:

  • To strengthen your basic knowledge and acquaint yourself with RPA technologies, we recommend completing the RPA developer foundation course from academy.uipath.com.
  • Learn basic development concepts and complete RPA Associate or the UiPath Certified Advanced RPA Developer (UiARD) certification. 
  • Try not to stress too much regarding your performance during the initial days and focus on learning from experience and practical learning. Practice more, explore UiPath Go’s reusable library, and be ready to consume automation. Sufficient hands-on experience can help you talk about development during an interview.
  • It is advisable to be active on the UiPath Forum as we get to learn from others and receive help with our queries. There are also several documents available to us to understand in depth about any product and use-cases, challenges, and hackathons available throughout the community forum. s.
  • Join community events. Hackathons are the best way to grow knowledge and learn from others, and we strongly recommend participating in every hackathon. 
  • Use developer best practices from an RPA perspective and apply them when giving practical demos in an interview or during work development. Having an RPA developer best practice skillset will always keep you one step ahead in your journey in the long run. It will help you to become a better solution architect and manage delivery going forward.  

RPA in 2022-24

Kadam says that the RPA industry is here to stay for a much longer run. However, with the growing digital transformation in every industry, it is becoming increasingly important to adapt to new technologies and bring modernisation. 

He adds, “I would say the major transformation factor for the RPA industry would be to adapt to IoT robots and Cloud technology. The era of RPA cloud technology has already begun, but there is a huge scope for business between 2022 and 2024. In addition, the RPA industry will also evolve to bring Intelligent Character Recognition (ICR), Optical Character Recognition (OCR), Natural Language Processing (NLP), and Natural Language Generation (NLG) to become a house of Intelligent Automation suite.

Mishra says that with Process Mining and Task Mining, businesses will discover and strengthen their automation funnel while expanding their digital transformation initiative. The no-code platform will show a strong footprint across the industry, the cognitive automation space will scale and strengthen itself with DU, AI Center, Form AI-like capabilities, and UiPath’s vision of a ‘Bot For Everyone’ will begin showing a true impact.

The period of 2022 to 2024 will be a whole new era for automation and the RPA industry, and we must keep ourselves up to date with its fast-paced developments.

More Great AIM Stories

Sreejani Bhattacharyya
I am a technology journalist at AIM. What gets me excited is deep-diving into new-age technologies and analysing how they impact us for the greater good. Reach me at sreejani.bhattacharyya@analyticsindiamag.com

Our Upcoming Events

Conference, in-person (Bangalore)
Machine Learning Developers Summit (MLDS) 2023
19-20th Jan, 2023

Conference, in-person (Bangalore)
Rising 2023 | Women in Tech Conference
16-17th Mar, 2023

Conference, in-person (Bangalore)
Data Engineering Summit (DES) 2023
27-28th Apr, 2023

Conference, in-person (Bangalore)
MachineCon 2023
23rd Jun, 2023

Conference, in-person (Bangalore)
Cypher 2023
20-22nd Sep, 2023

3 Ways to Join our Community

Whatsapp group

Discover special offers, top stories, upcoming events, and more.

Discord Server

Stay Connected with a larger ecosystem of data science and ML Professionals

Subscribe to our newsletter

Get the latest updates from AIM