RPA Product Capabilities 2020- AA vs UiPath

I published an article towards the end of 2019 comparing the product capabilities of the AA and UiPath, the two leading products in this hypergrowth RPA market. This was post the launch of A2019 analysing how significant was this move by AA, and looking at how that reflected upon their product capability offerings vis-a-vis the competition in this winner-takes-all market. I did that by applying my earlier published Intelligent Automation framework that presents a 2 by 2 matrix with the degree of technology adoption on one axis and the level of process transformation on the other. The framewok categorise automation efforts into 4 focus areas i.e. Task, Process, Data & Customer.

Taking the same framework as the basis, this article does a similar analysis of the product capabilities between AA and UiPath. It focuses on the key product enhancements made by these two vendors in these 4 capability areas in the year 2020 as presented in the tabular matrix below. It does not necessarily consider external capabilities coming from technology partner ecosystem, or existing capabilities from previous years or the ones that are re-branded/re-grouped. Capabilities with high impact are highlighted in green, whereas capabilities with medium to low impact are highlighted in yellow. I have plotted a "Product Capability Radar" as presented below the table. I have further summarised key takeaways from each category area when comparing these capabilities on their technical merits between the two vendors.

Task Based Automation (Employee Focused)

This is one of the most important product capability areas when considering RPA as this is what fundamentally defines the RPA market – task automation for enhancing employee productivity. AA stands out as clear winner in this category doubling down on their product strategy from last year with the launch of A2019 (a browser-based development tool from the cloud with greater accessibility and presented as a unified tool to foster collaboration between both business users and experience developers). Simplifying task automation with in-application executions, from applications such as Salesforce or Excel, as productivity packs is an impactful and very practical feature from AA.

Bringing to life the idea of a personal digital assistant is a significant step forward for the RPA industry as a whole, something I have written about extensively in the past as to bring out real robotics capabilities moving up from automation. Both AA and UiPath have launched their version of this capability as AARI and Assistant. The word AARI (if not SIRI) has a charm to it to resonate with the customers simply from a marketing messaging perspective. Although introduced as desktop interfaces, the real innovation we still await is to see the interactions with the digital assistant becoming more human-centric enabled via voice or chat interface- perhaps in 2021. But certainly, the ability to search, select, manage, assign and run task automations tailored to an individual’s needs and role in an organisation using a digital assistant is a key step forward in redefining the future of work.

UiPath added Task Capture capability that can be set up to capture, compare, and analyse the same tasks as done by multiple resources in an organisation to come up with optimal or fuller process maps and automation candidates. But, although anonymized, the whole idea to centrally monitor and compare employee task executions can be a tricky capability to implement widely- Robot police?

Will we see more concrete product innovations from UiPath in this space in 2021 aimed towards their vision of “A robot for every person”, or will AA continue to lead this space that will ultimately define the fate of this market?

Process Based Automation (Org Focused)

End to end process automation capability area has got the most attention in the year 2020 with all the buzzwords such as Hyperautomation, Digital Process Automation etc. Combining Process Identification with RPA tools has been a common theme across RPA vendors, with UiPath being the first mover in this area amongst the two and with AA catching up in 2020.

Moving to cloud as a platform and providing SaaS offerings is another significant capability that both vendors have looked to progress with- AA with better packaging and pricing options for different customer segments based on their size, UiPath with richer functionality options such as Apps & Action centre - does that make up for not having a fuller low-code browser based development tool to democratise automation for the business users?

UiPath has been attempting to take a lead in this category as a continued focus from the last year and to be seen as a more mature end-to-end automation lifecycle management (ALM) platform. AA has taken a more thoughtful and selective approach when it comes to building the capabilities in this area. UiPath's latest offering as Test Suite is another step in this direction offering solution to meet testing challenges in RPA implementations. But that is stretched further to provide Application and API testing capability as well. Does this position UiPath as an RPA tool or a Test Automation tool? This can confuse the buyers in the RPA market. The automated UI testing capabilities, by re-applying the core UiPath tool functionality, allow for component testing of the robot workflow branches and can benefit RPA customers in testing the design of robots (verification). But certainly, robots would still need functional testing to ensure they meet the process requirements (validation) as independent of the development tool, plus one cannot ignore other critical non-functional testing such as performance and operational acceptance that still needs doing to minimize potential risks to business as usual.

UiPath’s Auto-scaling of IT infrastructure is a much-needed feature to optimize run time infrastructure costs as per bot utilization which is important as more and more businesses look to scale RPA implementations. And so is the shared Object repository capability being added, something that is fundamental to UI Automation projects and has been there in other popular UI Automation tools such as QTP/UFT for more than a decade.

Private Bot Store is an important capability addition from AA and is carefully crafted to promote reuse and collaboration, and not as an ALM for automation pipeline management.

Knowledge Based Automation (Data Focused)

The knowledge-based automation capability area focused on unstructured data processing and AI capabilities is one that is most intensive from an R&D and investment standpoint, and often this leads into make or buy (partner) decisions from a product strategy perspective.

UiPath’s investments and product offerings in this area are excellent. UiPath’s AI Fabric stands out and the consistent life-cycle management theme here gives an edge and clear-cut value to the enterprise customers whether they are only looking to use what is available as pre-trained ML models or have deeper expertise to build their own custom models, or a hybrid. Providing composable framework for Document Understanding again provides a structured approach to cater for a wider range of use cases for enterprise customers looking to adopt AI and achieve practical outcomes for document processing.

AA also has made significant ongoing enhancements to their IQ Bot offering, and as well as the ML based AI sense recorder catching up with UiPath’s AI Computer vision capability from previous year.

UiPath also launched the Data Services feature in late 2020 to easily define, store and share data across automations as a no-code solution from the automation cloud. This could be a good feature to store and share transient data between different automations (i.e., data captured by one automation being reused by a following automation) especially for longer processes with automations running in a sequence. Transient because once the execution is complete such data is no more needed and could just be deleted. Another meaningful use case could be to store metadata or static data used by automations. But using such capability to fetch and centrally store data from multiple underlying legacy and enterprise applications can be a problematic idea to say the least. Data for business entities in different enterprise applications may have different representations and structures and building a joined up rationalized data view of business entities such as a customer, or a product is a significant enterprise architecture effort. Any such data extracted using RPA automations and stored as centrally can quickly loose its persistence if not imported or transformed correctly, or synchronised regularly, or shares common data structures under the hood with the underlying application’s system of records. A good reference example of how some of this can be achieved is with Microsoft dataverse. The idea to use such non-persistent data for serving customers by a contact centre agent as example could lead into potential business risks. Most of the popular contact centre applications such as Zendesk are already heavily invested into providing similar customer 360 capabilities.

Value Based Automation (Customer Focused)

Considering the key capability opportunities in this area, AA has a distinct advantage post the launch of A2019 product to provide infrastructure hosting support in other operating systems such as Linux, and as well as with a browser-based development tool to provide a cross browser/device/os anytime access to the business users.

In 2020, UiPath has forged technology partnership with a no-code/low-code Chatbot platform Druid to offer Chatbot integration with RPA automations. This allows building conversational or chat interactions that can then be used to trigger RPA automation.

And lastly to consider the latest product capability in UiPath as Apps (different from the feature last year to cater for human in the loop functionality being renamed to Automation Hub) which provide canvas style no code app development capabilities from the cloud platform. Apps can be single or multi-form and the UI controls can be linked with RPA automations to trigger executions. The (over) marketing around the feature can be confusing to explain the real intended use for this feature. If the idea is to trigger robots with simple form-based inputs, then this potentially can be better achieved with the Assistant or the chatbot integration. And if it is to build Multi-experience low code apps that are multi-channel, customer-facing, and enterprise-grade, then again the question is about the positioning and sophistication of such capability- Is it meant to be used for RPA automations or as standalone Low code app solution? Can you use a no-code canvas app as an enterprise grade Contact centre or ERP solution pulling non-persistent data collected in the data services as extracted from legacy applications that may not extend APIs? Time to replace those legacy apps perhaps!

The below bar chart summarises the relative positioning of AA and UiPath product capabilities against these 4 category areas-

--------------------------------------------------------------------------------

I hope you find this analysis interesting, please express your comments below.

Opinions expressed are solely my own and do not necessarily express the views or opinions of my employer.