The vGeek Blog

Introduction to RPA

The acronym RPA, which stands for “Robotic Process Automation”, identifies a relatively young type of technology that is becoming more and more popular across the IT industry. While RPA solutions have been available on the market for almost two decades now, their level of maturity has reached a point where they are now widely adopted in almost any business area.

But what is the purpose of RPA? To condense it in just a few lines: RPA is a set of technologies and tools that aims at multiplying the effectiveness of human workers by partnering them with a digital counterpart capable of automating or augmenting the execution of any type of workflow or business process.

This is usually achieved by “bots” (the digital workers) who can perform many tasks, reaching the highest level of complexity: from a simple bot that ingests a spreadsheet and extracts data from it, to an AI-driven one that takes mortgage application forms and decides if the loan can be granted or not. In both cases the human worker is not replaced, but rather aided by this digital helper, which takes the responsibility of doing the grunt work while the employee can focus on tasks that generate more value.

One might think that the widespread adoption of such a technology would bring heavy social impacts, up to leaving numbers of people without a job. Although some social implications will undoubtedly follow, we might consider also the positive side: the opportunity for smart workers to get away from repetitive and boring tasks. Employees using their intelligence to teach a system how to work for them, will feel, without a doubt, far more engaged and professionally fulfilled rather than simply being that system.

If we stop to consider it for a moment, this is what we, “old school” systems administrators, have already been doing for a while: moving from the old way of managing systems – manually clicking buttons inside UIs or typing one liner bash commands – to the new era of orchestrating infrastructures with lines of code and API calls.

The Application Anywhere approach

There are a few established vendors in the RPA industry and Automation Anywhere (AA) – the first ever RPA company presenting at a TFD event – has its own unique vision and execution strategy. One of the key points of AA’s approach is that RPA should be democratic, i.e. accessible, usable by and – most of all – beneficial to the whole organization adopting it.

Business users should be able to use and embrace RPA easily, up to the point they might feel confident to autonomously experiment with it and build their own small bots to automate their tasks.

Developers should have all the tools required to create complex bots that are capable of interconnecting different breeds of applications, even those not originally designed to dialogue with a human being but which require to have data piped one through another as part of complex workflows.

Finally, IT should be able to handle the infrastructure scalability issues that arise when the number of deployed bots increases over time.

These challenges are addressed by AA’s architecture. Without going too much in detail, we can think of a control plane (which also includes the tools to design the bots) and a bot execution layer. The control plane is what AA calls “Control Room”, which consists of specialized microservices running on one or more (Windows) servers backed by an Oracle or MSSQL DB. Control Room can run on-prem or on any public cloud and it is where bots are coded (Bot Creator), stored (Bot Repository) and scheduled for execution. Control Room has strong security features to ensure that access to bots and their triggering is tightly controlled to avoid accidental massive deployment to wrong systems or intentional tampering of code and execution of rogue bots.

The fact that Control Zone is built adopting microservices obviously makes possible and simpler to integrate bot management into DevOps practices, including working with CI/CD pipelines for continuous bot development and delivery.

Bots are run directly onto endpoints, typically Windows workstations or servers (Linux support is coming later this year), in either an attended or unattended way, depending on whether they are triggered and managed by a human or just scheduled for execution. This requires that a small footprint agent (the Bot Runner) is installed on the endpoint. When the bot execution is triggered, Control Room pushes the bot to the Bot Runner which executes it and deletes it immediately after the job is completed. Data is processed locally and it never leaves the endpoint. All activities are logged for governance and auditing purposes.

This attention to security and compliance is even more important when the control plane is in the public cloud and endpoints are on-prem workstations or even mobile laptops.

Integration with AI

One of the most interesting aspects of AA’s offering is that it comes out of the box with full integration with AI solutions to build upon RPA capabilities. AA comes with an open platform (IQ Bot) and the customer is free to use their internal AI solution or any 3^rd party one, with the objective of augmenting bot capabilities and allow them to learn and adapt their behaviors depending on the scenarios they will face. Think about bots being able to understand UI components, document structures, email contexts, speech etc. and act autonomously upon this: it will be immediately evident that AI is the ideal complement for bots, as they make together a more robust and powerful RPA solution.

Transforming the organization for RPA adoption

Adopting RPA requires an organizational transformation, in some ways similar to the one driven by cloud adoption: any organization embracing RPA will need full executive-level commitment and backing. There will a learning curve to be tackled by developers, business processes analysts and IT specialists but, most of all, there will be a need for full buy-in and active involvement of end users. After all, RPA will be there to help them to do a better job by radically changing their approach to work.

How does AA help with these challenges? AA offers programs, targeted at different audiences, tailored to simplify and expedite the adoption of their solution. The first recommended step for anyone willing to learn would be to enroll in the Automation Anywhere University program, where free e-learning courses are available to help business users, developers and IT operations staff understand and master AA’s solution. These courses not only teach how to go from zero to RPA-hero in a relatively limited time, but will also help obtain accreditations and certifications.

AA also offers a Community Edition of their software, clearly not suitable for enterprises, but perfect for small business, developers and students who want to learn RPA and how to build and administer bots.

To touch a theme very close to my heart: any technology is as strong as its user base advocacy is and AA seems to have got this right by hosting an online community called A-people where developers and users connect and collaborate to improve their skills and work together on bot-building.

Finally, to get users started and be operational from scratch, ready-to-use bots are available in the Bot Store, a concept very similar to the app marketplaces for our beloved mobile devices. While most of the available bots are free to download and use, AA is launching also a monetization program, so that developers can create, publish and support paid-for bots.

Final thoughts

I must admit that, when I first heard about RPA, I was both skeptical and worried, especially for the social implications of this technology. Attending Automation Anywhere’s session at TFD 19 has been eye-opening: my concerns were based on my little knowledge of the purpose and the objectives of RPA solutions. Now that I have a wider and deeper understanding of this kind of technology I have realized that it is not here to steal jobs from people, but to make their time at the desk matter more, while at the same time empowering organizations to provide higher quality services (and more of them) at a fraction of the cost. RPA is here to stay and it is only at the beginning of its journey. Let’s meet again in five years and discuss about its impact on everyone’s lives.

Disclaimer: travel, hotel and meal expenses for my participation to Cloud Field Day 4 have been kindly paid for by Gestalt IT, who invited me as a delegate. I am not under any obligation by neither Gestalt IT nor any of the vendors who participated to TFD19 to write any review or recommend any of the products and solutions presented at the event. I have not received any compensation for writing the above post and its content only represent my personal opinions.

If you like this post, please share it!