RPA and AI Meetup (1): The Path to AI
This is the sixth RPA and AI meetup hosted by Bloomsearch, introduced by Will Kellett, MD of Bloomsearch we had 3 presentations interlinked around digital transformation and the advances in AI. Once again we ran out of beer though, that's disappointing but the sign of a good time!
The Path to AI
Andrew Rayner, VP of Customer Success at UiPath.
"UiPath strategy is to build an ecosystem of technologies"
UiPath were co-hosting the event and the session was opened by Andrew Rayner, a 2 year UiPath veteran, who formerly held the role of product development before moving over to customer success.
UiPath is one of the fastest growing software companies in history, with $400m investment and 2500 enterprise customers. They were also the first RPA company to offer free training to address the growing need for skills and are being generally disruptive in a market they helped to create in the first place.
Andrew began the session with perhaps one of the clearest explanations of the different components which make up a typical RPA setup. I'm certain we've all seen the standard definitions, which look something like
Studio - where the developer works to create the bots
Orchestrator - which controls the work
Robots (Runtime) - who do the work
Andrew compared this to music, where the composers write the music (Studio), the conductor organises the performance (Orchestrator) and the individual musicians play the music (Robots). If you're ever stuck explaining why there are different elements, this is the best description I've ever heard.
UiPath have taken a deliberate approach to make their environment open and easy to integrate with. The path to AI of the title is really describing how many of these integrations are AI, extending the standard rote transaction processing of traditional RPA and adding the ability to learn new skills. Andrew described four elements to the path to AI:
This involves the reliable recognition of objects in images and on the user interface, converting semi-structured pictures and images into structured data that can be utilised by intelligent OCR (optical character recognition). UiPath already has a track record with computer vision involving integrations with Microsoft, Google and ABBYY; each of which can be used to identify the type of document and determine the Next Best Action.
The classification and extraction of data from documents based on historical transactions is the area Machine Learning is most recognised. Solutions such as ABBYY Flexicapture templates and Celeton, Skymind and Data Robot were highlighted as examples of ecosystem suppliers in this area.
Opening up the use of Machine Learning for those more technically minded, it is possible to run python scripts directly from within UiPath. The ecosystem continues to expand, meeting new business needs through the use of ML.
Natural language interactions through chatbots or voice are changing the ways we interact with computer systems, this is the ability for intelligent identification and information extraction of unstructured (voice of language) into structured data for processing.
Process monitoring (or Process Mining) helps to rapidly uncover what is happening. Where is the work? how long is it taking? where are the transactions? Tools such as Mineit or Celonis are finding the opportunities
The strategy to build 'soft partnerships' has clearly been successful with a wide range of 3rd party solutions mentioned as part of the UiPath ecosystem. There is clearly an intent to further strengthen these partnerships with the training academy including these products in the offering.
The integration with AWS, Azure for cloud, a range of workflow management tools and chatbot integrations and developer friendly solutions such as cyberark demonstrate success to date.
To close the presentation Andrew reflected that mergers and acquisitions may be on the horizon to meet demand but the principle of an open and extensible platform will remain their core focus.
Look out for future articles from the other two presentations from the London RPA and AI Meetup. These are:
A Strategy for Intelligent Automation: Thomas Puschnik, Group Head of Intelligent Automation at Zurich Insurance Group.
The Explainability of AI: Derek Dempsey, Director of Data Science at FICO.
Rob is the Author of Digital Workforce, co-founder of Wzard Innovation Ltd, VP Product at the RPA Academy and associate of the Camelot Round Table Consultants.