1. Peter Schooff
  2. BPM Discussions
  3. Tuesday, 05 December 2017
  4.  Subscribe via email
What sort of impact do you see AI having on processes in the year ahead?
Ian Gotts Accepted Answer Pending Moderation
AI can be huge a force multiplier for well designed processes. Think of AI as automation on steroids.

But like automation, AI applied to poor, inefficient or badly designed processes will simply add costs and build a case for the AI hype cycle.
  1. one week ago
  2. BPM Discussions
  3. # 1
Karl Walter Keirstead Accepted Answer Pending Moderation
Walk . .. then run.

I hope the impact will be minimal.

Unless/ until operational staff across multiple industries/applications get to where they are gaining both efficiency and effectiveness from in-line BPM in run-time Case Management platforms, no need to try to skip over less-complex methods/techniques that can have a dramatic impact on building and sustaining competitive advantage.

AI is great for highly automated processes, not so great for work that is a mix of structured and unstructured work.
Oh yeah.. Give it all to AI and run away fast, not to be hit too hard!
  1. Boris Zinchenko
  2. 1 week ago
  1. one week ago
  2. BPM Discussions
  3. # 2
Bogdan Nafornita Accepted Answer Pending Moderation
For the immediate future, the successful AI applications around pattern recognition problem classes will enhance user interfaces, accelerating data input and consumption.

Not much outside of that.
Managing Founder, profluo.com
  1. one week ago
  2. BPM Discussions
  3. # 3
John Morris Accepted Answer Pending Moderation
How big will the impact of AI be on "processes"? Meaning, if I'm a business executive, should I count on AI projects taking up valuable resources this coming year in order to put us ahead? Here's a model for thinking about AI adoption (which applies beyond just process-oriented projects):

1. PARADOX OF AI BUSINESS CASES: "The best AI use cases are not necessarily the best AI business cases." The best business cases for AI involve huge risk or cost reductions. But the very nature of these business cases means that business likely already has a non-AI solution in place (via older technology and management processes). So, now the AI business case is not from zero, but for replacement. So "wait and see" is relatively viable. However -- see next item . . .

2. GENUINE NEW USE CASES: "Genuine new AI use cases imply business disruption, or they don't matter". There are genuine new AI use cases -- in other words situations where AI can make a big difference in risk or cost, but where older technologies weren't feasible or economically viable. By definition, if such a use case exists and especially given the criteria of "big difference", then such a use case will drive industry disruption at scale. Genuine AI use cases, if they are economically significant, mean one has to get ready to disrupt one's business model. If you can't afford to support disruption at scale, or alternatively if your nice new AI use case isn't that important, "wait and see" is relatively viable. However -- see next item . . .

3. PRODUCTIZATION AND COMMODITIZATION: "Broad AI adoption will only happen when you can buy it off the shelf". Because otherwise it's too expensive for most organizations, technically and concerning skill levels. If I can buy a tunable image recognition system to deploy for some small but nagging problem, I'll do it. I won't hire a data scientist for edge cases. I'll let the vendor hire the data scientist. And when I hire the "tunable image recognition system", I probably won't think of it as AI. We are beginning to see such products come to market. Maybe we won't have to "wait and see".

As AI moves thru it's current "Hype Cycle" ( read "social signalling" ), we see the unfortunately named "Trough of Disillusionment" ahead. That only means time to get to work on engineering and business analysis and productization. Any given business executive knows their business domain. What are the AI use cases and business cases that can be made? This is the time of incremental steps. And because business processes are necessarily at the heart of all work, AI use cases and business cases will be intimately enmeshed with business process infrastructures. The business purpose of AI is to support decisioning, probably at scale. And repetitive decisioning is encoded in organizational business processes. AI and BPM therefore go hand-in-hand. AI will have an impact on your processes in 2018 if you make it happen.
RE "The best AI business cases are not the best business cases." - reminded me "The best engineering solutions are not necessary the best political solutions".
+1 @Alexander, nice comparison. After thinking about it I changed the original above to a comparison of "best use cases versus best business cases" . . .
  1. John Morris
  2. 1 week ago
  1. one week ago
  2. BPM Discussions
  3. # 4
Patrick Lujan Accepted Answer Pending Moderation
Blog Writer
Very little. Should really work on mastering, optimizing the technological capabilities to-date, and how best to leverage them on behalf of the business, and quit being distracted by every new technical bauble, shiny potential panacea that comes along.
  1. one week ago
  2. BPM Discussions
  3. # 5
Siva Puvvada Accepted Answer Pending Moderation
Most of the use cases will be around recommendation engines that drive better customer experience.

Examples include recommending best options to a customer, selecting a best task an employee can work on or deciding between a robot worker and a human worker.

Another application will be around OCR that will help streamline document oriented processes.
  1. one week ago
  2. BPM Discussions
  3. # 6
David Chassels Accepted Answer Pending Moderation
I think AI may evolve in 3 aspects of business; machines, intelligent processes and intelligent UIs
Briefly machines or gadgets with clever programming etc take in data and create an output ready for use. Intelligent Processes can assimilate data from previous actions in the work environment and dynamically create appropriate next required actions. The intelligent UIs recognise the user and can dynamically present data that is required based upon the user's input. Yes some will be seen as Automation but whatever all are surrounded by people and processes and important this fundamental point must be recognised before any form of AI is used. Just adopting AI without understanding and being connected to the end to end process could be costly mistake. However hopefully the rise of BPM will put all into perspective to deliver some significant benefits to all users.
  1. one week ago
  2. BPM Discussions
  3. # 7
Kay Winkler Accepted Answer Pending Moderation
AI wont necessarily make too much of a difference in the BPM market during 2018, but most definitively over the next couple of (5?) years. However, I do expect it be very granular in nature, where AI would actually apply in BPM. While I don't see too much of a use case and added benefit of AI, orchestrating the logical flow of a process, AI as part of a self learning business rules engine that operates on a form level, very much could be a game changer. I would a "business AI" expect to be a service that evolves and enriches BPMS in parallel - but not actually becoming part of the suite (eg. a business process that maintains several connections to Watson, on a subscription based model).
Bigger vendors such as IBM, of course, will be able to offer access to AI as an integral and native part of their respective BPM, ECM, CCM and BI solution stacks.
NSI Soluciones - ABPMP PTY
DIsagree about big vendors such as IBM having advantage. Next generation no low code platforms will intuitive support creation and use of AI in any work place process and UI environment. AI machines will be likely supplied by specialist vendors and readily available for use as appropriate and supported by surrounding BPM supported software.
  1. David Chassels
  2. 1 week ago
  1. one week ago
  2. BPM Discussions
  3. # 8
Pedro Robledo Accepted Answer Pending Moderation
It should be necessary to think about what AI? I mean, not all AI disciplines will be presented next year in the same way, but great value for BPM: 1.)Cognitive Systems will be important in terms than they are able to use Natural Language to process the communication between machine and human, so processes can use this systems to automate this communication, and these systems can take decisions when they understand what are the customer asking for, and decide what action has to be made to solve the question. 2.) Machine Learning will be important for RPA (Remote Process Automation), so it is possible to automate tasks using RPA once Machine Learning has learned the consecutive steps to repeat in the best way to solve the activity in the process. 3) As BigData will be very important for any company who wants to have more competitiveness, they will require Deep Learning to create neural networks to transform BigData in Knowledge to be processed.

And structured processes are not always possible to have, as the process change every time that occurs so it is require to define in real time, so what about Process Adaptability, you have to think about Adaptability Case Management. In this arena, many BPM experts we have collaborated in the annual book INTELLIGENT ADAPTABILITY of Future Strategies to write about the application of AI in Case Management to apply in unstructrured processes (a sure trend in the next year), and the result has been published in November, so if you want to know more about this, you can read the book. Intelligent Adaptability describes how ACM is emerging in the era of machine intelligence and automation technologies, including Big Data, digitization, Internet of Things (IoT), artificial intelligence (AI), intelligent BPMS and BPM Everywhere. In my article “Smart Adaptive BPM Engine”, I write in the book about my vision about the future of the BPM Suites, as disruption from Artificial Intelligence applied to BPM will provide a smart adaptive BPM engine that will cause fundamental changes for the structured processes because these could change their predesigned behavior as the BPM engine starts learning, so this new capability of BPM will provide the right tasks at the right moment for a customer. (If you are interested I provide the link to my post about the book)
  1. https://pedrorobledobpm.blogspot.com.es/2017/11/new-book-intelligent-adaptability-just.html
@Pedro - great book!!
@David - thanks for sharing your thoughts. I really hope that's the way it will play out. It certainly would democratize the play-field to an important degree. I do have concerns however, how "real" of an AI smaller providers would bring to the table, actually. You see, I often think that nowadays the AI term is mostly misused somewhat, representing complex (or even simple) yet set algorithms. In my mind, AI starts with independent, cognitive abilities of programs, actually coming up with their own algorithms, based upon a given pattern recognition. There, I guess, bigger players have had more time and budget available for more mature and capable solutions. But that's just my personal reasoning, based on very little detailed insight and tangible data to back that up :D - So, you may be completely right. I also think that a pluralistic vendor scenario, where smaller but specialized players have an important role, is more favorable for the market overall, anyways.
  1. Kay Winkler
  2. 1 week ago
  1. one week ago
  2. BPM Discussions
  3. # 9
Pritiman Panda Accepted Answer Pending Moderation
Artificial Intelligence is a multidimensional subject area. It can be broadly classified as a blend of Machine Learning, Predictive/Adaptive Analytics, NLP, Text Analytics, Voice Pattern Recognition, Image Analytics, Deep Learning, Graph Analysis, Robotics and many more.
The current state for adoption of AI complementing Business Processes in an enterprise is a bit chaotic with Discussions, Debate, PoCs, R&D etc. Hopefully, the coming year will embark on a more structured & focused approach(fingers crossed)
From process impact standpoint, the key topic that will make a difference few years down the line are:
  • Enhanced Decisioning Strategies
  • ‎Predictive / Adaptive Analytics
  • Persona Driven solutions
  • ‎Mature Robotics Process Automation
  • ‎Development of Self Healing Processes
  • ‎Inception of ML & Deep Learning Model driven processes
  • ‎Improvised Customer Experience (personalized) with Data fuelled strategies
  • ‎Reduction of OpEx with automated decisioning & self-healing capabilities
  • ‎Hype & marketing around "touchless" or "zero-touch" implementation e.g Touchless Claim Processing, Zero Touch Onboarding etc.
  • ‎Reduction of Contact Center overheads with Chatbots/Virtual Assistant implementations
  • ‎The AI, IoT & Process interlock can also act as a game changer in certain scenarios
  • ‎BPM products adding yet another capability to their stack as a new feature (be it homegrown or a plug & play model) :-)

To summarize:
AI implementation is not new for enterprise (processes), they are already doing it in some form or the other (magnitudes may differ). But yes with the buzz in the wild, it has become a checklist mandate and a parameter for competitive advantage. Definitely, with new tools / technologies mushrooming everyday, AI will strengthen/mature & proliferate to unlock tremendous value. Things won't change overnight to attain the "nirvana" state, it will be a steady incremental approach with a judicious feasibility study.

In a nutshell, irrespective of "getting Robots out of the humans[RPA]" or "humanizing the Robots…..by adding more intelligence[AI/Cognitive]" - the "process" still remains the heart/pulse and the enterprise ecosystem. An inefficient or weak process design is enough to choke the arteries, give cramps or cardiac arrest & dismantle the Humanized Bot

[Artificial Intelligence]:[Brain of the Enterprise] :: [Business Process]:[Heart of the Enterprise]
Boris Zinchenko Accepted Answer Pending Moderation
BPM, at least in respect to process automation and process mining, is indispensable part of AI. Impact of modern AI methodologies is continually increasing over past years. BPM is, to considerable extent, a projection of AI into business process world.

Role of BPM for AI is equally important. Explosive growth of AI in recent years goes on substantial extent chaotically. Important business transactions delegated to AI come completely outside of control of company management. BPM is the only safety belt capable of protecting companies from their new born artificial brains and retaining control in AI driven digital platforms. Not accidentally, all leading AI systems increasingly use BPM techniques and methodology as an essential part of their business applications.

Example: Diagram of GoogleNet architecture

  1. https://research.googleblog.com/2017/05/using-machine-learning-to-explore.html
@Boris' comment is compelling. I interpret it as follows: The question about AI and BPM is now reframed from "possible impact of AI on process" to a state-of-the-union such that (1) "BPM is one business projection of AI" and (2) "BPM is a strategic interface to AI". In other words, an association between BPM process technology and AI is not opportunistic, but necessary and important. This association is especially important for one of the most important challenges around AI, which is the provenance and audit trail of any decision (i.e. the fear that machines are making decisions about which we have little to no idea concerning how those decisions have been taken). BPM technology is about modeling and managing the work of business, explicitly. Thus naturally insofar as AI helps us with our work, we would expect the two technologies to be closely associated. SUMMARY: Work on both AI and BPM at the same time -- or else.
  1. John Morris
  2. 3 days ago
  • Page :
  • 1

There are no replies made for this post yet.
However, you are not allowed to reply to this post.