≡ Menu

Jim Sinur of Aragon Research recently published a new blog Mounting Pressure for Better Decisions. He argues, correctly, that decision making is under pressure because there is more data available than ever before, a need for faster change in the way organizations make decisions to respond to evolving circumstances and a general need for speed in handling transactions.

We help companies improve decision-making by applying our DecisionsFirst Decision Management approach and by building Decision Management Systems for them. Combining decision models (built using the Decision Model and Notation or DMN standard) with powerful business rules management systems, advanced analytics (machine learning, predictive analytics) and AI, we help companies see a set of unique benefits:

  • Improved consistency
    Decision models enable consistent decision making across channels and people without imposing mindless consistency.
  • Increased Agility
    The systems we build are easy for the business to change in response to new business conditions because the business understand the decision models and own the business rules that drive the system.
  • Reduced Latency
    The combination of business rules and advanced analytics enables higher rates of straight through processing (automation) while also ensuring more clarity and less confusion for the transactions that must be handled manually.
  • Lower Cost
    Decision Management Systems reduce costs by ensuring less waste and rework, more STP and fewer manual touches.
  • Better Accuracy
    Decision Management Systems operationalize data-driven, analytical decisions throughout the organization to improve the accuracy of decisions everywhere.

If you are interested in learning more about Decision Management and the technology available for it, check out our Decision Management Systems Platform Technology Report or contact us for a free consultation.

Maureen Fleming of IDC presented at IDC Directions on How Does Decision-Centric Computing Drive Digital Transformation? She kindly shared this presentation with me. Decision-centric computing, she says:

continuously receives and analyzes data to predict when decisions need to be made, systematically learns how to automate those decisions, and acts on each decision to improve performance.

Exactly. We call these Decision Management Systems but the concept is the same.

While the presentation focused on IoT and streaming scenarios, the concepts can be applied more generally – after all, many business scenarios are heading to a streaming solution. The most interesting piece was this graph titled “Predictive Analytics is Only a Piece of the Puzzle”
This graph shows that organizations that are mature in terms of predictive analytics use business rules a lot (70%), those that are in production with something use business rules a little (24%) and those that are stuck in development are not using them very much at all (5%).

This illustrates a point we make with analytics clients – a business rules management system is a great platform for deploying predictive analytics, especially when you apply Decision Management principles and decision modeling to do the rules in a decisions first way.

For IDC subscribers, Maureen has written Introducing Decision-Centric Computing which has another great quote:

Without a way to incorporate decision automation to make repetitive decisions, enterprises will find it increasingly difficult to justify their investments in advanced analytics and risk failure to materialize the anticipated benefits

Decision Management is a proven approach to delivering Decision-Centric Computing and using a Decisions First methodology effectively combines business rules and predictive analytics using decision modeling. What are you waiting for?

Jim Sinur, VP of Research and Aragon Fellow at Aragon Research recently posted “Better Decisions with Decision Management” to his blog. Jim begins by describing Decision Management as “another discipline that will help consistently deliver better decisions”, especially when added to analytics and AI.

It’s great to have Jim’s focus turn to a Decision Management Framework and a Decision Management Platform – we are excited to see what he comes up with.

Of course we use Decision Management on all our projects, applying our unique Decisions First approach to ensure success. Check out the Decision Management Manifesto for our philosophy and if you want our take on a Decision Management Platform, check out the Decision Management Systems Platform Technologies Report with lots of detail on current technology and approaches, use cases and more.

Analytics are only valuable if your enterprise’s decision making changes for the better. You need to build an analytic enterprise that leverages analytics to inform strategy, empower people, and especially drive systems.  An analytic enterprise uses analytics to solve its most critical run-the-business problems, and uses increasingly advanced and diverse analytics to maximize its ability to get value from data.

There are three critical success factors for building an analytic enterprise -focusing on business decisions, moving to predictive and prescriptive analytics and focusing on continuous improvement not one-time big wins. You can learn more about how and why to become an analytic enterprise in this white paper Building An Analytic Enterprise and the associated webinar recording here.

This research was sponsored by Teradata.

Customers, IBM says, are moving to the cloud but they are transitioning through a hybrid solution. IBM is investing heavily in its cloud in terms of partnerships, technology, patents, volume, data centers etc. They announced two new partnerships this week – Cloudlfare and New Relic.

The One Cloud architecture is particularly focused on AI and analytics enablement – cloud infrastructure that assumes you want to use the data on the cloud to drive analytics and AI. It’s also very API-centric and designed to be managed programmatically. Plus the Watson APIs are fully integrated along with the various data capabilities IBM has been developing for its cloud.

IBM Cloud Private is IBM’s key platform for modernizing applications. They are adding capabilities around application transformation, developer tools and the data cloud. Integration across multiple clouds and deployment automation /management are focus areas also.

Transformation Advisor scans existing applications to assess the complexity of migrating an existing application to a container-based environment. If possible it will automate the transformation to containers. Once containerized the IBM Cloud Private catalog allows these applications and standard ones to be deployed to multiple instances and provides monitoring for them once deployed. Applications can also be pushed to public clouds and monitored there also. Plus of course there’s a command line interface for all this.

All good stuff. Of course you should also think about replacing all that hard-wired code with decision-centric business rules too….

 

Building a data-driven culture, where evidence-based decisions support bottom-line business objectives and AI is embedded into workflows across your organization. Ensure data is secure and accessible, wherever it lives, and get insights from data and turn them into competitive advantage. Use the entire spectrum of data science, artificial intelligence and machine learning to lay a foundation for a fast-approaching future where AI isn’t just an advantage, it’s essential.

Rob Thomas, GM Analytics for IBM, kicked off a session on putting data to work with AI. Rob began talking about the impact standard shipping containers had on the shipping industry and how a similar move is required in data – something that will make it easy to combine and analyze data in a standard way. And that only this kind of data landscape can support systematic application of analytics and AI.

ING came on stage to talk about their information architecture – one that addresses regulatory issues but also makes it possible for everyone to access, understand and use data for better decisions. They pulled all their data into a data lake architecture and then mapped the core of this to a standard set of corporate data models/vocabulary based on industry models. On to this they layered governance etc. Plus this supports the application of AI both to improve the data and its metadata AND to improve decision-making.

IBM has a new solution offering – IBM Cloud Private for Data. This is designed to provide an out of the box environment for managing an organization’s data and supporting its broad an deep application of AI and analytics. It makes it easy to bring on-premise and cloud data, tracks machine learning models running against the data and provides integrated search and preview across the metadata for all this data.

Beth Smith came on stage to add Watson and AI into this mix. Lots of organizations lack the AI skills they need so IBM is launching IBM Watson Studio to help AI teams collaborate around the data an organization has, working easily with the new IBM Cloud Private for Data. It’s open, supporting open source as well as IBM-specific AI capabilities like the pre-trained Watson APIs. It’s underpinned by a catalog that combines data and any analytics you have built against it. It also supports and automates many of the experimentation and training runs that good ML and AI models require – helping reduce the manual load on data scientists – while providing a rich visual interface for much of the work. It’s designed to make it easier to build, easier to run and easier to share the tasks needed for AI.

IBM has also been investing in the services support that companies need and launching the Data Science Elite Team to deliver initial free workshops to help companies get over the hump and get started with more sophisticated analytics and AI.

Nice to see the investment in making AI and analytics easier. Wish IBM would include its Business Rules Management System Operational Decision Manager as part of this stack – would make operationalizing the result much easier.

Ginni Rometty kicks off the main event with her opening keynote focusing on putting smart to work. Her premise is that everything could be changing now because business and technology architectures are changing at the same time – something that does not happen very often. The opportunity is for exponential change across all businesses thanks to the combination of data and AI. And she further argues that the fact that so much of the data that is needed is INSIDE companies makes it possible for established companies to compete – to disrupt and not just be disrupted.

Digital Platforms are the key to this. She emphasized that multiple platforms are going to matter -no-one is going to use just one. These platforms will allow you to embed intelligence in every process across the organization.  She feels that AI is going to be in combination with people when it is used most effectively. And she encourages companies to go on offense – to use this intelligence to  not just fix things but to really grow exponentially. Plus IBM’s business model is not to monetize their clients data but to help their clients do so.

Social disruption is possible too – everyone needs to focus on trust, jobs/skills and inclusion. If AI is a complement to human intelligence then IBM thinks that all jobs will be disrupted – some will be eliminated, some will be created, all will be changed.

Lots of announcements coming she says around cloud, especially making it easy to integrate private cloud into public ones, around strategic partners, and around Watson, especially around making it easier to use Watson and embed it in work.

Customers up next to help reinforce Ginny’s points. Verizon CEO first talking about 5G, about strategic partners in the API economy. In particular they want to build better ecosystems around their core transmission capability. He also emphasized the importance of data management and trust, especially for a network. Key point – building a platform but partnering to build things on top of it.

Maersk – one of the world’s largest container companies – came up next to talk about how they worked with IBM to use blockchain to disrupt the way shipping works. Shipping companies are coming on board to digitize the way they share information about containers and vessels/vehicles. Using blockchain to make it easier to share and update information in a trusted way. And the organizations participating include government agencies, insurance, ports and much more. A good example of the value of making an open platform not just a company one.

RBC – Royal Bank of Canada – came up next. One of the biggest changes, the CEO says, is the way people look for the financial services they need – they go online where before they would have come to a bank branch. Mobile and internet payment platforms mean that people don’t see the brand any more – they set the card up in an app once. And mobile is changing the way they run their back office systems. All of this puts pressure on their ability to develop everything – especially AI – so they are partnering and moving to cloud. And of course because its money you can’t just push something out there and see if it works – people want it fast but they want it secure and reliable too. In particular, RBC sees using AI to really improve customer service and customer engagement.

Ginny came back to re-emphasize that this an inflection point as simultaneous business architecture and technology architecture changes create a once in a lifetime opportunity to become “an incumbent disrupter”.

DecisionCAMP 2018 in in Europe – Luxembourg to be precise – September 17-19. This is a great event and well worth your time if you are interested in the nuts and bolts of decisioning technology, Decision Management or decision modeling. Last year’s event in London was great with a wide range of presentations and lots of great content. Plus you get to meet with a bunch of folks really committed to decision-making approaches and technologies.

Anyway, its time to submit papers – the Call for Papers is here. If you have something to say about decision modeling, the use of business rules and analytic or AI technology for decision automation, optimization, how decision management and blockchain can deliver smart contracts, or really anything else interesting and decision-centric, please go ahead and send a proposal. Like the rest of the committee I am looking forward to seeing some great topics again this year.

Get those submissions in by March 25 if you can – or at least let us know you plan to!

The Decision Management Community is trying to establish a most influential people list for Decision Management specifically. The plan is to have members of the community vote based on nominations provided here. So if you have someone you think has demonstrated leadership, engagement and innovation in the Decision Management community, why not go ahead and nominate them? And if you are not already a member of the DM Community, why not register so you can vote and stay in touch with the articles and news the site collates.

AI is a decision-making technology. A focus on decisions, not a separate AI initiative, delivers business value and a strong ROI.

A recent HBS survey of executives adopting Artificial Intelligence (AI) provides critical context for companies considering how best to invest in AI:

  • Few companies have made much progress to date — most are experimenting. You still have time to consider how best to invest in AI.
  • AI works best in companies that have already invested in digitizing their business as it enhances digital channels, digital decisions and digital processes.
  • While there is plenty of hype, AI works when it is implemented correctly.

As companies invest in AI technologies, it is clear a technology-led approach does not work. To get business value from AI, companies should focus AI efforts on improving business decisions. We have just published a new brief that lays out a clear, straightforward approach to succeeding with AI by leading with business decisions. It can be applied if you have not yet begun or to focus and reset efforts that aren’t making the progress you desire.

Get the brief here.

OneClick.ai is a company taking advantage of the fact that many AI problems use similar approaches to reduce the time and cost of individual AI projects. It was founded and received its initial funding in 2017, and launched the product last year. The company has a core team of 8 in the US and China with 40 active enterprise accounts supporting over 20,000 models.

OneClick.ai uses AI to build AI and so help companies get into AI more quickly and more cheaply. The intent is to get them fault-tolerant scalable APIs for custom-built AI solutions in days or even hours instead of weeks and months. They aim to automate the end-to-end development of AI solutions based on deep learning. They use meta-learning to design and evaluate millions of deep learning models to find the best ones. They are also working on capabilities to explain how those models work, to address one of the concerns of deep learning, the lack of interpretability.

The product is aimed at non-technical users with a chatbot interface to allow experts to interact with the trained models. Users can choose from public cloud, private cloud or hosted versions and software vendors have the access to an OEM version to integrate the technology into customized solutions. A wide range of AI use cases are supported, including classic predictions (weekly and monthly sales or equipment failure) to image recognition (recognize brands in shelf images to see how much shelf space they have), classification (putting complaint emails into existing categories and identifying new problems) and semantic search (find the most helpful supporting material for a fault). Several of their existing customers were already trying to use AI and have found OneClick.ai significantly quicker to get to an accurate model.

The tool is browser-based and supports multiple projects. Each project has a chatbot that can answer data science questions. Data is provided by uploading flat files that contain a learning data set – numeric, categorical, date/time, text or images. Raw data is enough but users can add domain-specific features if they have domain knowledge that a feature will likely be helpful. Users can develop classification, regression, time-series forecasts, recommendations or clustering models and target various measures of precision depending on the type of model – accuracy, mean absolute error etc.

The engine builds many models and presents the best from which the user can select the one they prefer (based on their preferred metric and the latency of the deployed model, which is calculated for each model). The engine automatically keeps 20% out for testing and uses the other 80% for training. Under the covers, the engine keeps refining the techniques it uses based on the previous training results. Once built the chatbot can answer various questions about the models such as usage tips and model comparison. Users can deploy the models as an API for real-time access with few clicks. A future update will also allow model updates and deployment through an SDK.

You can find out more here.

March 27-29 I am teaching a 3-part online live training class that will prepare you to be immediately effective in a modern, collaborative and DMN standards-based approach to decision modeling.

ExampleYou’ll learn how to identify and prioritize the decisions that drive your business, see how to analyze and model these decisions, and understand the role these decisions play in delivering more powerful information systems.

Each step in the class is supported by interactive decision modeling work sessions focused on problems that reinforce key points. All the decision modeling and notation in the class conforms to the DMN standard, future-proofing your investment in decision modeling. DMN-based decision modeling works for business rules projects using a BRMS, predictive analytic or data science projects, manual or automated decisions and even AI.

Click here for more information and registration. Early bird pricing is available through March 1, 2018 so book now!

I have written before on how a decisions-first approach is ideal for success with AI. After reading David Roe‘s article 11 Questions Organizations need to Ask Before Buying into AI I thought a few more comments were in order:

If you focus on decisions first and on how you must/could/want to make the decision, you can rapidly tell if you really need AI at all. Often business rules and simpler analytics are enough – but you need to know what decision you are trying to make before you can tell. Similarly if you don’t know what else, besides the AI, is going into the decision then you won’t be able to tell how much impact AI is going to have. It’s easy to have a compliance or policy constraint undermine the “lift” you get from AI.

The business case for most AI is “better decisions”. If you don’t know which decisions, and what counts as better, then your AI is just a gimmick. Know what decisions you are trying to improve and how before you begin to ensure your AI has a real business case.

Decision models are great for showing you what else goes into a decision besides AI. This let’s you see how exposed you are when the AI gets it wrong, how good your predictions need to be to be helpful and much more. Understand the context first and it’s easier to manage, and get support for, your AI plans.

Lastly integrate AI into your decisioning stack -make sure your business rules, predictive analytics, machine learning and AI can be integrated to deliver a single, better decision (based on a decision model).

If you want to learn more about decision modeling, contact us or come to our live online decision modeling with DMN training in March.

 

Back in November I posted a humorous Thanksgiving guest decision model to LinkedIn. I just repeated the exercise with a decision model to help you assess a New Year’s Resolution.

While these are just for fun, I thought it might be worth sharing how I built this one. Normally we like to work top-down talking to business experts but in this case I did not have any to work with so I had to start bottom-up with research.

  1. I started with some articles – found using google – and each became a Knowledge Source in the diagram.
  2. I looked over each and identified the things it implied you should decide about a New Year’s resolution to help you decide if it was a good one or not.
  3. As I added these Decisions to my model, I connected them to the Knowledge Sources that related to them (some Decisions recurred in several articles, of course).
  4. One set of Decisions – deciding if a resolution met the five criteria to be SMART – could be grouped as sub-decisions of a higher-level Decision.
  5. Others were grouped based on thematic elements – a common approach where there is not a specific structure driven by regulation or similar.
  6. This gave me a structure – the Decision I was trying to model, some high-level sub-decisions and logically grouped sub-decisions.
  7. Cleaning up the diagram required putting copies of the Knowledge Sources on the diagram (though these point to the same instance in the underlying repository).

In this case I didn’t deal with Input Data as the model seemed useful with just Decisions and Knowledge Sources. To finish it, we would need to identify data elements and write decision logic (or develop predictive models) for each element.

If you are interested in decision modeling, why not register for our upcoming live online Decision Modeling with DMN class in March.

Happy New Year.

SAS Decision Manager is SAS’ platform for decision automation and is getting a significant update in December 2017. I wrote a product review of SAS Decision Manager in 2014 and a number of things have changed in the new release, which is on the new SAS Platform and leverages new SAS Viya technologies.

SAS Decision Manager is aimed at an analytics ecosystem that is a moving target these days with cloud-enabled analytics that are more open and API-driven,  more people doing data science, and different kinds of data coming to the fore. Meanwhile IoT is adding new data streams and demanding decision-making at the edge while machine learning and AI are hot trends and offer real possibilities.

“If analytics does not lead to more informed decisions and more effective actions, then why do it at all”
Mike Gualtierei, Forrester.

This quote embodies the need to operationalize these analytics and enable faster decision making. SAS believes, as we do, that one must put analytics into action, operationalize your analytics, to get value. You need to go from data to discovery and to deployment. In this context, SAS Decision Management is a Portfolio to create and manage decisions:

Overall architectural view

  • SAS Model Manager – import and govern models, monitor and retrain models, deploy models. And increasingly any kind of models including R, Python…
  • SAS Decision Manager – build business rules, build decisions that use analytics and rules in a decision flow, deploy as decision services. The SAS Business Rules Manager product has been subsumed into the new SAS Decision Manager product to create a single environment.
  • SAS Event Stream Processing Studio – SAS Event Stream Processing Studio is now in the SAS Decision Management portfolio so that decisions can be injected into the streaming data environment – real time as micro services but also into streams.
  • Execution – covers Cloud Analytic Service (Viya) for testing and deployment as well as model training, Micro Analytic Service for REST ESP for streaming data, and in-database or in-Hadoop.
  • Plus, open APIs to allow REST, Python, Lua, Java and CLIs to access the platform. R and PMML can be brought into the modeling tools too.

SAS Decision Manager wraps business rules, analytic models, flow logic (and soon Python) into services while linking to Model Manager to access the models being used. These models are developed in the new SAS VDMML Model Studio. The new release of SAS Decision Manager is built on the new SAS Platform, which brings the benefits of the new platform around cloud readiness, multi-tenant etc. This release also combines the Business Rules Management offering in SAS Decision Management.

Key elements overall include:

  • Visual Decision Modeling – decision simulation and path tracing, model and business rule integration and streamlined business rules management
  • Unified publishing to ESP, Cloud Analytic or Micro Analytic services, in-database or in-hadoop
  • Model Manager integration to make it easier to share models and support for more kinds of models as well as managing publishing of models to multiple end points (e.g. in IoT) and automating updates etc.
  • Open APIs from Viya, workflow etc.

Some specific improvements for SAS Model Manager

  • Common Model Repository with GUI and REST interfaces to manage content and search to find the right models
  • Can register models from SAS VDMML Model Studio and import models from PMML, Python, Zip files, etc.
  • Model publishing to various defined targets from in-DB, In-Hadoop, SAS, streaming or real-time with SAS micro analytic service.
  • Model compare in terms of statistics and plots as well as the definition of champion and challenger.
  • Version control with revert, tracking, creation of new versions

SAS Decision Manager

  • Decision inventory in a common repository with access to the models in the model repository as well as to the rules available. All these elements are versioned.
  • New graphical decision flow editor that brings analytic models from model manager, rules and specific branching logic.
  • The testing environment shows how data flows through the decision flow to show which paths were most heavily used. Data can be brought in dynamically or from existing data sets.
  • New editor allows direct access to the model or rules from the flow and get access to repository information as the diagram is edited. Rules are managed directly in the same repository
  • Can create temporary information items on the fly for use in rules
  • Can bring in lookup tables from the SAS data environment
  • Ruleset editor allows data to be pulled in as the vocabulary (copying from another or accessing the data layer) and then rules can be written.

Test data results showing which elements of the decision flow have the most transactions.

In addition to the December released, the plan is to move to more regularly update the product with a 6 month cycle for with new algorithms, more integrations, more use of the Viya APIs etc.

You can get more information on SAS Decision Manager here.

All industry standards offer interchange. Successful standards offer skills interchange not just a technical interchange format.

The Decision Model and Notation (DMN) decision modeling standard has a published XML interchange format, of course, and several of the committee’s members are working really hard to iron out the remaining issues and make the XML interchange more robust. The ability to interchange decision models between vendors is a valuable one but the opportunity that DMN offers for skills interchange is, if anything, even more valuable.

DMN offers two critical kinds of skills interchange – it offers those working with business rules or decision logic a way to transfer their skills between products and it offers business analysts a way to transfer skills between different kinds of decisioning projects.

The vast majority of the business logic in a decisioning system can be defined using the two core DMN components:

  • Decision Requirements Diagrams structure decision problems, break them into coherent pieces. They show where data is used and what knowledge assets (policies, regulations, best practices) are involved.
  • Decision tables specify the logic for most of the decisions on the diagram using simple constructs.

You don’t get 100% of the execution defined using these two elements -you need to add “glue” of various kinds – but almost 100% of the business content is defined using these them. This means someone who knows DMN can transfer these skills between DMN tools. But it also means they can transfer these skills between business rules products too as the approach of decomposing a decision problem into a Decision Requirements Diagram before writing logic is totally transferable and frankly most decision tables look and work the same even if they don’t support DMN yet.

The second kind of skills interchange comes because decision modeling works for lots of different kinds of projects. We have used decision modeling and DMN to:

  • Define business rules / decision logic for automation
  • Frame requirements for predictive analytics and machine learning project
  • Orchestrate a mix of packaged and custom decisioning components including business rules,  predictive analytics, AI and optimization
  • Model manual decision-making for consistency, mixing manual and automated decision making
  • And more – see Decision Modeling has value across many projects

This means that business analysts who learn decision modeling can apply that skill across lots of projects.

Learn decision modeling and learn DMN. It’s a great skill that let’s you express business decision problems and one that is transferable -interchangable – across projects and products.

BPMInstitute.org would like to get your insights on how you’re using Digital Decisioning and Analytics in your organization. Your feedback will help shape articles and focus at BPM Institute for 2018.

Digital Decisioning and Analytics survey

  • Are you using analytics and reporting to innovate business functions and models?
  • What is the state of your analytics efforts as they relate to processes?

Share your insights with BPMInstitute.org and you’ll be entered into a random drawing to win one free OnDemand course of your choice

 

 

AI is a hot topic and we get asked a lot by clients how they can succeed with AI or cognitive technology. There’s often a sense of panic – “everyone is doing AI and we’re not!” – and a sense that they have to start a completely separate initiative, throw money at it and hope for the best. In fact, we tell them, they have some time – they need to keep calm and focus on decisions.

The folks over at HBR had a good article about adopting AI based on a survey of executives. This is well worth a read and makes a couple of critical points.

  1. AI really does work, if you use it right. There’s plenty of hype but also plenty of evidence that it works. But like all technologies it works when it works, it’s not a silver bullet.
  2. Not everyone is using AI – in fact hardly anyone is doing very much with it. Most regular companies are experimenting with it, trying it out in one small area. Despite what you read there’s still time to figure out how to use AI effectively in your organization. Stay Calm.
  3. AI works better if you have already digitized your business. Of course AI is a decision-making technology, so what matters here is that you have digitized decision-making.  Focus AI on digital decision-making.

To succeed with AI we have a concrete set of suggestions we give to customers, many of which overlap with the HBR recommendations as you would expect:

  • Get management support
    The best way to do this is to know which decisions you are targeting and show your executives how these decisions impact business results. Being able to describe how improving a particular decision will help an executive meet their objectives and exceed their metrics will get their attention.
  • DON’T put technologists in charge
    Like data analytics, mixed teams work best for AI. Make sure the team has business, operations, technology and analytics professionals from day 1. For maximum effectiveness, use decision modeling with DMN to describe the decision-making you plan to improve as this gives everyone a shared vision of the project expressed in non-technical terms.
  • Focus on the decision not AI
    You will want to mix and match AI with other analytic approaches, explicit rules-based approaches and people-based approaches to making decisions. Most business decisions involve a mix:

    • Rules express the regulatory and policy-based parts of your decision
    • Data analytics turn (mostly) structured data into probabilities and classifications to improve the accuracy of your decisions
    • People make the decisions that involve interaction with the real world and poorly scoped or defined ones
    • And AI handles natural language, image processing, really complicated pattern matching etc.
  • Make sure you focus on change management
    Change is always a big deal in Decision Management projects – as soon as you start changing how decisions are made and how much automation there is you need to plan for and manage change. AI is no exception – it will change roles and responsibilities and change management will be essential for actual deployment (distinct from a fun experiment).

AI is a decision-making technology. As such it is a powerful complement to Decision Management – something to be considered alongside business rules and analytics, and integrated into a coherent decision model. Here’s one example, for a company that needed to automate assignment of emails. This depended on who it was from, what it was about and how urgent it was:

  • Deciding which client an email was from involved rules run against the sender and sender’s domain.
  • Deciding on the subject of an email involved rules about senders (some automated emails always use the same sender for the same subject) and rules about subject lines (some are fixed format).
  • This left too many unclassified, however, so the subject and body of the text were analyzed using text analytics to see which products were mentioned in the email to identify them (analytically) as the subject of the email.
  • Urgency was hard too. Historical data about the client was analyzed to build customer retention model. This analytic score was used to increase the urgency of any email from a client who was a retention risk.
  • Finally AI was used to see what the tone of the email was – was the email a complaint or a problem or just a description? The more likely it was to be a problem or complaint, the higher the urgency.
  • Each of these sub-decisions used different technologies but were orchestrated in a single decision model to decide how to assign the email.

AI is certainly new and different, but success with it requires the same focus on decisions and decision-making. Put decisions first.

My friends at Actico recently had me record some videos on Decision Management and decision modeling with DMN. Here’s the first – 3 reasons why financial and insurance companies should adopt Decision Management.

Enjoy.

As part of the build up to Building Business Capability 2017 I gave an interview on transforming the business. Check it out.

If you want to come to BBC 2017, there’s still time to register with code SPKDMS for a 10% discount.

If you are coming to BBC 2017, don’t forget to register for my tutorial Decision-Centric Business Transformation: Decision Modeling. See you there.