Embeddable AI saves time building powerful AI applications

Just a few weeks ago, IBM announced an expansion to their embeddable AI software portfolio with the release of three containerized Watson libraries. This expansion allows our partners to embed popular IBM Watson capabilities, including natural language processing, speech-to-text, and text-to-speech into their applications and solutions. But what is embeddable AI, and what are its uses?

Embeddable AI is the first-of-a-kind suite of IBM core AI technologies that can be easily embedded within enterprise applications to serve a variety of use cases. Think of embeddable AI as an engine. Planes and cars both use engines to make them go, but each engine accomplishes its purpose in different ways.

The analogy doesn’t end there either, because just like with a car or plane engine, it’s much easier to use something pre-made than to construct one yourself. With embeddable AI, you get a set of flexible, fit-for-purpose AI models that developers can use to provide enhanced end-user experiences—like, automatically transcribing voice messages and video conferences to text.

Unfortunately, many organizations still struggle to find talent with the skills to build and deploy AI solutions into their businesses. So, it’s crucial for companies that want to add specific AI capabilities to their applications or workflows to do so without expanding their technology stack, hiring more data science talent, or investing in expensive supercomputing resources.

To address this, businesses have found value in embedding powerful technology using specific models to harness AI’s potential in the way that best fits their needs, whether it is through domain optimized applications to containerized software libraries.

Differentiated solutions drive business success

IBM Research has added three new software libraries to IBM’s portfolio of embeddable AI solutions—software libraries that are not bound to any platform and can be run across environments, including public clouds, on-premises, and at the edge.

As a result, organizations can now use this technology to enhance their current applications or build their own solutions.

The new libraries include:

◉ IBM Watson Natural Language Processing Library for Embed

◉ IBM Watson Speech-to-Text Library for Embed

◉ IBM Watson Text-to-Speech Library for Embed

In addition to the libraries, the embeddable AI portfolio includes IBM Watson APIs and applications like IBM Watson Assistant, IBM Watson Discovery, IBM Instana Observability, and IBM Maximo Visual Inspection.

Embeddable AI libraries are lightweight and provide stable APIs for use across models, making it easier for organizations to bring novel solutions to market.

Partner solutions using embeddable AI

IBM partners are making use of embeddable AI in various ways and across different industries.

LegalMation, an IBM partner that helps the legal industry make use of AI and advanced technology, uses natural language processing to automate contract privacy. Contracts and agreements contain information that organizations want to be careful about. Usually, redacting information within a contract is a manual process involving a person going line by line to mark passages for redaction. Instead, LegalMation uses embeddable AI to create an automated solution. The legal company now uses a natural language processing tool to find and mark sensitive information automatically.

See how LegalMation also uses AI to reduce the early-phase response documentation drafting process from 6 – 10 hours to 2 minutes.

Language-training school ASTEX, based in Madrid, Spain, has seen student careers skyrocket after they completed its courses. ASTEX uses AI to streamline students’ onboarding experience, offer personalized learning plans and improve the program’s scalability by reducing its dependence on humans. IBM partner Ivory Soluciones connected ASTEX with IBM because of the tech company’s expertise in AI solutions. Working closely together, ASTEX, Ivory, and IBM developed the ASTEX Language Innovation platform on IBM Cloud® with IBM Watson® technology.

Call recording service, Dubber, uses speech-to-text, tone analyzer, and natural language understanding to capture and transcribe a variety of verbal exchanges. The solution, powered by embeddable AI, automatically translates phone calls and video conferences into text and assigns each conversation a positive, negative, or neutral value, depending on the call. Users can then mine the data using simple keyword searches to find the information they need.

Now, with the addition of the new software libraries, new and existing IBM partners can embed the same Watson AI that powers IBM’s market leading products with flexibility to build and deploy on any cloud in the containerized environment of choice.

Source: ibm.com

Continue reading

3 key reasons why your organization needs Responsible AI

Responsibility is a learned behavior. Over time we connect the dots, understanding the need to meet societal expectations, comply with rules and laws, and to respect the rights of others. We see the link between responsibility, accountability and subsequent rewards. When we act responsibly, the rewards are positive; when we don’t, we can face negative consequences including fines, loss of trust or status, and even confinement. Adherence to responsible artificial intelligence (AI) standards follows similar tenants.

Gartner predicts that the market for artificial intelligence (AI) software will reach almost $134.8 billion by 2025. 

Achieving Responsible AI

As building and scaling AI models for your organization becomes more business critical, achieving Responsible AI (RAI) should be considered a highly relevant topic. There is a growing need to proactively drive responsible, fair, and ethical decisions, designed to:

Manage risk and reputation

No organization wants to be in the news for the wrong reasons, and recently there have been a lot of stories in the press regarding issues of unfair, unexplainable, or biased AI. Organizations need to protect individuals’ privacy and build trust. Incorrect or biased use of AI based on faulty datasets or assumptions can result in lawsuits and an erosion of stakeholder, customer, stockholder and employee trust. Ultimately, this can lead to reputational damage, lost sales and decreased revenue.

Read More: C1000-059: IBM AI Enterprise Workflow V1 Data Science Specialist

Adhere to ethical principles

The importance of driving ethical decisions – not favoring one group over another, requires AI systems that achieve fairness. This necessitates the detection of bias during data acquisition, building, training, deploying and monitoring models.  Fair decisions require the ability to adjust to changes in behavioral patterns and profiles. This may demand model retraining or rebuilding.

Protect and scale against government regulations

AI regulations are growing and changing at a rapid pace and noncompliance can lead to costly audits, fines and negative press. Global organizations with branches in multiple countries are challenged to meet local and country specific rules and regulations. Organizations in highly regulated markets such as healthcare, government and financial services have additional challenges in meeting industry regulations around data and models.

“The average cost of compliance came in at $5.47 million, while the average cost of non-compliance was $14.82 million. The average cost of non-compliance has risen more than 45% in 10 years. The true cost of non-compliance for organizations due to a single non-compliance event is an average of $4 million in revenue.”  The True Cost of Noncompliance

Responsible AI requires governance

Despite good intentions and evolving technologies, achieving responsible AI can be challenging.  AI requires AI governance, not after the fact but baked into AI strategy of your organization. So what is AI governance? It is the process of defining policies and establishing accountability to guide the creation and deployment of AI systems.

For many of today’s organizations today, governing AI requires a lot of manual work that include  the use of multiple tools, applications and platforms. Lack of automation can lead to lengthy model approval, validation and deployment cycles during which model drift and bias can happen. Manual processes can lead to “black box models” that lack transparent and explainable analytic results.

Explainable results are crucial when facing questions on the performance of AI algorithms and models. Your company’s management, stakeholders and stockholders expect accountability.  Your customers deserve and are holding your organization accountable to explain reasons for analytics-based decisions. These may include credit, mortgage and school denials, or the details of healthcare diagnosis or treatment. Documented, explainable model facts are necessary when defending analytic decisions.

An AI Governance solution driving responsible, transparent and explainable AI workflows

The right AI governance solution can help to better direct, manage and monitory your organization’s AI activities. With the right end-to-end automated platform, your organization can strengthen the ability to meet regulatory requirements, protect the reputation of your organization and address ethical concerns.

The IBM AI Governance solution automates across the AI lifecycle from data collection, model building, deploying and monitoring. Model facts are centralized for AI transparency and explainability. This comprehensive solution comes without the excessive costs of switching from your current data science platform. This solution includes:

Components of the solution include:

Lifecycle governance

Monitor, catalog and govern AI models from where they reside. Automate the capture of model metadata and increase predictive accuracy to identify how AI is used and where models need to be reworked.

Risk management

Automate model facts and workflows for compliance to business standards. identify, manage, monitory and report on risk and compliance at scale. Dynamic dashboards provide customizable results for your stakeholders and enhance collaboration across multiple regions and geographies.

Regulatory compliance

Translate external AI regulations into policies for automated enforcement. This results in enhanced adherence to regulations for audit and compliance purposes and provides customized reporting to key stakeholders.

Source: ibm.com

Continue reading

Four starting points to transform your organization into a data-driven enterprise

Due to the convergence of events in the data analytics and AI landscape, many organizations are at an inflection point. Regardless of size, industry or geographical location, the sprawl of data across disparate environments, increase in velocity of data and the explosion of data volumes has resulted in complex data infrastructures for most enterprises. Furthermore, a global effort to create new data privacy laws, and the increased attention on biases in AI models, has resulted in convoluted business processes for getting data to users. How do business leaders navigate this new data and AI ecosystem and make their company a data-driven organization? The solution is a data fabric.

A data fabric architecture elevates the value of enterprise data by providing the right data, at the right time, regardless of where it resides.  To simplify the process of becoming data-driven with a data fabric, we are focusing on the four most common entry points we see with data fabric journeys. In 2023, we have four entry points aligned to common data & AI stakeholder challenges.

We are also introducing IBM Cloud Park for Data Express. These are solutions that are aligned to the data fabric entry points. IBM Cloud Pak for Data Express solutions provide new clients with affordable and high impact capabilities to expeditiously explore and validate the path to become a data-driven enterprise. IBM Cloud Pak for Data Express solutions offer clients a simple on ramp to start realizing the business value of a modern architecture.

Data governance

The data governance capability of a data fabric focuses on the collection, management and automation of an organization’s data. The automated metadata generation is essential to turn a manual process into one that is better controlled. In this way it helps avoid human error and tags data so that policy enforcement can be achieved at the point of access rather than individual repositories.  This data-driven approach makes it easier to find the data that best fits their needs of business users. More importantly, this capability enables business users to quickly and easily find the quality data that conforms to regulatory requirements. IBM’s data governance capability enables the enforcement of policies at runtime anywhere, in essence “policies that move with the data”. This capability will provide data users with visibility into origin, transformations, and destination of data as it is used to build products.  The result is more useful data for decision-making, less hassle and better compliance.

Data integration

The rapid growth of data continues to proceed unabated and is now accompanied by not only the issue of siloed data but a plethora of different repositories across numerous clouds. The reasoning is simple and well-justified with the exception of data silos; more data allows the opportunity to provide more accurate data-driven insights, while using multiple clouds helps avoid vendor lock-in and allows data to be stored where it best fits. The challenge, of course, is the added complexity of data management that hinders the actual use of that data for better decisions, analysis and AI.

As part of a data fabric, IBM’s data integration capability creates a roadmap that helps organizations connect data from disparate data sources, build data pipelines, remediate data issues, enrich data quality, and deliver integrated data to multicloud platforms. From there, it can be easily accessed via dashboards by data consumers or those building into a data product. The kind of digital transformation that an organization gets with data integration ensures that the right data can be delivered to the right person at the right time. With IBM’s data integration portfolio, you are not locked into just a single integration style. You can select a hybrid integration strategy that aligns with your organization’s business strategy to meet the needs of your data consumers wanting to access and utilize the data.

Data science and MLOps

AI is no longer experimental. These technologies are becoming mainstream across industries and are proving key drivers of enterprise innovation and growth, leading to more accurate, quicker strategic decisions. When AI is done right, enterprises are seeing increased revenues, improved customer experiences and faster time-to-market, all of which leads to revenue gains and improvements in their competitive positioning.

The data science and MLOps capability provides data science tools and solutions that enable enterprises to accelerate AI-driven innovation, simplify the MLOps lifecycle, and run any AI model with a flexible deployment. With this capability, not only can data-driven companies operationalize data science models on any cloud while instilling trust in AI outcomes, but they are also in a position to improve the ability to manage and govern the AI lifecycle to optimize business decisions with prescriptive analytics.

AI governance

Artificial intelligence (AI) is no longer a choice. Adoption is imperative to beat the competition, release innovative products and services, better meet customer expectations, reduce risk and fraud, and drive profitability. However, successful AI is not guaranteed and does not always come easy. AI initiatives require governance, compliance with corporate and ethical principles, laws and regulations.

A data fabric addresses the need for AI governance by providing capabilities to direct, manage and monitor the AI activities of an organization. AI governance is not just a “nice to have”. It is an integral part of an organization adopting a data-driven culture. It is critical to avoid audits, hefty fines or damage to the organization’s reputation. The IBM AI governance solution provides automated tools and processes enabling an organization to direct, manage and monitor across the AI lifecycle.

IBM Cloud Pak for Data Express solutions

As previously mentioned, we now provide a simple, lightweight, and fast means of validating the value of a data fabric. Through the IBM Cloud Pak for Data Express solutions, you can leverage data governance, ELT Pushdown, or data science and MLOps capabilities to quickly evaluate the ability to better utilize data by simplifying data access and facilitating self-service data consumption. In addition, our comprehensive AI Governance solution complements the data science & MLOps express offering. Rapidly experience the benefits of a data fabric architecture in a platform solution that makes all data available to drive business outcomes.

Source: ibm.com

Continue reading

Using a digital self-serve experience to accelerate and scale partner innovation with IBM embeddable AI

IBM has invested $1 billion into our partner ecosystem. We want to ensure that partners like you have the resources to build your business and develop software for your customers using IBM’s industry-defining hybrid cloud and AI platform. Together, we build and sell powerful solutions that elevate our clients’ businesses through digital transformation.

To that end, IBM recently announced a set of embeddable AI libraries that empower partners to create new AI solutions. In fact, IBM supports an easy and fast way to embed and adopt IBM AI technologies through the new Digital Self-Serve Co-Create Experience (DSCE).

The Build Lab team created the DSCE to complement its high-touch engagement process and provide a digital self-service experience that scales to tens of thousands of Independent Software Vendors (ISVs) adopting IBM’s embeddable AI. Using the DSCE self-serve portal, partners can discover and try the recently launched IBM embeddable AI portfolio of IBM Watson Libraries, IBM Watson APIs, and IBM applications at their own pace and on their schedule. In addition, DSCE’s digitally guided experience enables partners to effortlessly package and deploy their software at scale.

Your on-ramp to embeddable AI from IBM

The IBM Build Lab team collaborates with qualified ISVs to build Proofs of Experience (PoX) demonstrating the value of combining the best of IBM Hybrid Cloud and AI technology to create innovative solutions and deliver unique market value.

DSCE is a wizard-driven experience. Users respond to contextual questions and get suggested prescriptive assets, education, and trials while rapidly integrating IBM technology into products. Rather than manually searching IBM websites and repositories for potentially relevant information and resources, DSCE does the legwork for you, providing assets, education, and trial resources based on your development intent. The DSCE guided path directs you to reference architectures, tutorials, best practices, boilerplate code, and interactive sandboxes for a customized roadmap with assets and education to speed your adoption of IBM AI.

Embark on a task-based journey

DSCE works seamlessly for both data scientist and machine learning operations (ML-Ops) engineers’ personas.

For example, data scientist, Miles wants to customize an emotion classification model to discover what makes customers happiest. His startup provides analysis of customer feedback to help the retail e-commerce customers it serves. He wants to provide high-quality analysis of the most satisfied customers, so he chooses a Watson NLP emotion classification model that he can fine-tune using an algorithm that predicts ‘happiness’ with greater confidence than pre-trained models. This type of modeling can all be done in just a few simple clicks:

◉ Go to https://dsce.ibm.com ->

◉ Find and try AI ->

◉ Build with AI Libraries ->

◉ Build with Watson NLP ->

◉ Emotion classification ->

◉ Library and container ->

◉ Custom train the model ->

◉ Results Page

The bookmarkable Results Page gives a comprehensive set of assets for both training and deploying a model. For accomplishing the task of “Training the Model,” Miles can explore interactive demos, reserve a Watson Studio environment, copy a snippet from a Jupyter notebook, and much more.

If Miles, or his ML-Ops counterpart, Leena, wants to “Deploy the Model,” they can get access to the trial license and container of the new Watson NLP Library for 180 days. From there it’s easy to package and deploy the solution on Kubernetes, Red Hat OpenShift, AWS Fargate, or IBM Code Engine. It’s that simple!

Try embeddable AI now

Try the experience here: https://dsce.ibm.com/ and accelerate your AI-enabled innovation now. DSCE will be extended to include more IBM embeddable offerings, satisfying modern developer preferences for digital and self-serve experiences, while helping thousands of ISVs innovate rapidly and concurrently. If you want to provide any feedback on the experience, get in touch through the “Contact us” link on your customized results page.

Source: ibm.com

Continue reading

Do you need coding skills to build a chatbot?

Virtual assistants are valuable, transformative business tools that offer a compelling ROI while improving the customer experience. By 2030, conversational AI chatbots and virtual assistants will handle 30% of interactions that would have otherwise been handled by a human agent—up from 2% in 2022. So why haven’t more companies adopted this solution?

Due to the maturity and complexity of the technology, it can take years to reap the full benefits of developing a conversational AI platform. One challenge to implementing conversational AI from core Natural Language AI technology is that it requires expensive technical specialists. These specialists (data analytics, AI and graph technologies) are often in short supply and typically require annual salaries of $175,000 or more. But for most companies there is a simpler, cost-effective way to create a chatbot for the ideal support experience.  

A multitude of solutions to create and maintain virtual assistants to improve customer service.  

Keep in mind that not all pre-built tools are the same. Historically, conversational AI solutions have fallen into one of two categories:  

Simple to use solutions

These tools make it easy to start working with conversational AI but in the end offer sub-par customer experiences. Typically, the answers are hardcoded or use minimal AI or machine learning. These chatbots are programmed to give exact answers to specific questions — if client says this, then say this, if client says that, then say this. 10 times out of 10 they will give the same answer regardless of whether it is right or wrong.  

Robust solutions that can create powerful experiences

On the other end of the spectrum are these tools which allow you to create the type of experiences your customers expect, but they’re often very complex (and expensive) to use. For instance, they may have robust natural language processing powering their AI but to build a customer facing solution requires a degree in computer science and a front-end developer to create the experience and embed it into your website. 

Watson Assistant changes the game with actions and the low code/no code interface  

You may feel limited by expensive AI solutions that require a wide variety of tools and technologies to build, integrate with existing systems, and operate at scale to meet peek customer demand. If your organization is not equipped to develop conversational AI applications to handle dialog and process and understand natural language, consider a Conversation AI platform with “low code” and “no code” interfaces that line-of-business (LOB) users can use to quickly develop conversational AI applications that are ready for your enterprise.

Breaking the code: Get started fast with a Conversational AI platform 

Watson Assistant’s new no-code visual chatbot builder focuses on using actions to build customer conversations. It’s simple enough for anyone to build a virtual assistant, with a build guide that is tailored to the people who interact with customers daily.  

Pre-built conversation flows

First, we focus on the logical steps to complete the action, with AI that is powerful enough to understand the intent, recognize specific pieces of information (entities) from a message and keep the user on track, all without being redundant. For example, a customer may write, “I want to buy one large cheese pizza.” Should the assistant ask then “What size pizza?” “What toppings?” and “How many?” No, it should just skip to the payment. Watson Assistant AI will understand that if you start on part 3, parts 1 and 2 should be skipped.  

Disambiguation

It’s simple to design a virtual assistant that asks clarifying questions. If you say something in the chat and the virtual assistant doesn’t understand it, it will automatically ask clarifying questions to get the conversation back on track. Which means that developing your bot doesn’t have to be perfect right out of the box. It will automatically get the conversation back on the rails. You can also have predetermined responses. The chatbot could ask: “What is your account?” and if it’s not getting a number, it will automatically clarify and say, “I’m looking for a number only, please.”

Maintenance

With Watson Assistant, you can identify and address any chatbot problems in a matter of minutes, as opposed to the traditional development cycle.  This allows you to avoid the wasted time and resources associated with going to IT, assigning a developer, and waiting weeks or even months before the changes are actually made. 

Low-code and no-code interfaces like Watson Assistant’s visual chatbot builder go beyond professional developers and seasoned technologists to open up a whole new class of citizen developers. Companies can take back control to quickly and easily build chatbots for customer service, no code needed. 

Source: ibm.com

Continue reading

Call Center Modernization with AI

Picture this: A traveler sets off on a camping trip. She decides to extend her RV rental halfway through her trip, so she calls customer service for assistance, but finds herself waiting minutes, then what feels like hours. When she finally does get a hold of somebody, her call is redirected. More waiting follows. Suddenly her new plan doesn’t seem worth the aggravation. Now, imagine the same scenario from the agent’s perspective, dealing with a dissatisfied customer, scrambling for information that takes time to collect. Instances like these are far too common—the debacle ends up being costly for the company, and frustrating for both customer and agent.

Conversational AI solutions for customer service have come a long way, helping organizations meet customer expectations while reducing containment rates, complexity, and costs. It starts with bringing AI into the mix and ends with more cost-efficient operations and more satisfied customers.

So how can conversational AI help fulfill customer expectations in today’s ever-demanding landscape?

When you deploy conversational AI in your call center, you get:

1. Increased customer and agent satisfaction. Think of the example above—long wait times and unanswered questions can only lead to frustrated customers and agents and slower businesses. With leading natural language understanding (NLU) and automation leading to faster resolution, everybody wins.

2. Improved call resolution rates. AI and machine learning enable more self-service answers and actions and help route customers who need live agent support to the right place – continuously analyzing customer interactions to improve response. Agents benefit from this assistance too; empowering them to perform at their best when call traffic is high. Ultimately, improved resolution rates mean better customer experiences and improved brand reputation.

3. Reduced operational costs. With the capabilities of AI-powered virtual agents, you can contain up to 70% of calls without any human interaction and save an estimated USD 5.50 per contained call. This is money saved for your business, and time saved for your customers.

Not all AI platforms are built the same

On the lowest rung of the AI ladder, you have rules-based bots with limited response function. For example, you want to know if your telecom provider offers an unlimited data plan, so you call customer service and are given a set of basic questions following strict if-then scenarios—“…say yes if you want to review service plans; say yes if you want unlimited data.”

Climb up one rung, and there’s level two AI with machine learning and intent detection. You accidentally type “speal to an agenr”— but the virtual assistant understands your intention and responds properly: “I will connect you with an agent who can assist you.”

Then there’s IBM Watson® Assistant—the always-learning, highly resourceful virtual agent. Watson Assistant sits at the top—level three. Level three offers powerful AI that has unparalleled data and research capabilities.

The Watson Assistant deployed at Vodafone, the second-largest telecommunications company in Germany, exhibits level-three capacities—in addition to answering questions across a variety of platforms, such as WhatsApp, Facebook and RCS, Watson Assistant answers requests pulled from databases and can converse in multiple languages. It mines data, customizes interactions and is continuously learning. “*Insert Name*, transferring you to one of our agents who can answer your question about coverage abroad.” 

With Watson AI, you can expect more for your call center: 24/7 support, speedy response times and higher resolution rates. Seamlessly integrate your virtual agent with your existing back-end systems and processes, with every customer channel and touchpoint, without migrating your tech stack—IBM can meet you wherever you are in your customer service journey. Watson AI offers:

◉ Best-in-class NLU

◉ Intent detection

◉ Large language models

◉ Unsupervised learning

◉ Advanced analytics

◉ AI-powered agent assist

◉ Easy integration with existing systems

◉ Consulting services

All these features work in concert to redefine customer care at the speed of your business.

Why add complexity when you can simplify with AI? 

According to a Gartner® report, in 2031, conversational AI chatbots and virtual assistants will handle 30% of interactions that would have otherwise been handled by a human agent, up from 2% in 2022. To remain among the leaders, modern contact centers will need to keep up with AI innovations. Of course, like Watson, leading businesses are constantly learning, analyzing, and striving to become better.

Watson Assistant plugs into your company’s infrastructure, is reliable, easy to use, and always there to provide answers and self-service actions. Take Arvee, for example, an IBM Watson AI-powered virtual assistant for Camping World, the number one retailer of RVs. When customer demand surged early in the global pandemic, Camping World deployed Arvee in their call center and agent efficiency increased 33%.  Customer engagement also increased by 40%.

Similarly, IBM is working together with CcaaS providers like Nice to make it even simpler to build, deploy and scale AI-powered virtual voice agents.

Watson Assistant helps streamline processes and create agent efficiency—and when calls go to human agents, they can deliver higher quality personal service. Remember that aggravated customer from earlier? With the power and capabilities of Watson Assistant, she can enjoy her time camping—goodbye hold music, hello sounds of nature.

Source: ibm.com

Continue reading

How data, AI and automation can transform the enterprise

Today’s data leaders are expected to make organizations run more efficiently, improve business value, and foster innovation. Their role has expanded from providing business intelligence to management, to ensuring high-quality data is accessible and useful across the enterprise. In other words, they must ensure that data strategy aligns to business strategy. Only from this foundation can data leaders foster a data-driven culture, where the entire organization is empowered to take advantage of automation and AI technologies to improve ROI. These areas can transform the enterprise, from cost savings to revenue growth to opening new business opportunities.

Building the foundation: data architecture

Collecting, organizing, managing, and storing data is a complex challenge. A fit-for-purpose data architecture underpins effective data-driven organizations. Driven by business requirements, it establishes how data flows through the ecosystem from collection to processing to consumption. Modern cloud-based data architectures support high availability, scalability and portability; intelligent workflows, analytics and real-time integration; and connection to legacy applications via standard APIs. Your choice of data architecture can have a huge impact on your organization’s revenue and efficiencies, and the costs of getting it wrong can potentially be substantial.

The right data architecture can allow organizations to balance cost and simplicity and reduce data storage expenses, while making it easy for data scientists and line of business users to access trusted data. It can help eliminate siloes and integrate complex combinations of enterprise systems and applications to take advantage of existing and planned investments. And to increase your return on AI and automation investments, organizations should consider automated processes, methodologies, and tools that manage an organization’s use of AI through AI governance.

Taking advantage of automation for LOB and IT activities

You can use data to completely digitize your organization with automation and AI. The challenge is bringing it all together and implementing it across lines of business and IT.

For line-of-business functions, here are five key capabilities to consider:

1. Process mining to identify the best candidates for automation and scale your automation initiatives before investments are carried out

2. Robotic process automation (RPA) to automate manual, time-consuming tasks

3. A workflow engine to automate digital workflows

4. Operational decision management to analyze, automate, and govern rules-based business decisions

5. Content management to manage the growing volume of enterprise content that’s required to run your business and support decisions

6. Document processing to read your documents, extract data, and refine and store the data for use

Looking at the digitization of IT, here are three capability areas to evaluate:

1. Enterprise observability to improve application performance monitoring and accelerate CI/CD pipelines

2. Application resource management to proactively deliver the most efficient compute, storage, and network resources to your applications

3. AI to proactively identify potential risks or outage warning signs across IT environments

Help increase ROI on data, AI and automation investments by making data and AI ethics a part of your culture

But process and people can’t be ignored. If you don’t properly infuse AI into a major process in an organization, there may be no real impact. You should consider infusing AI into supply chain procurement, marketing, sales, and finance processes, and adapt processes accordingly. And since people run the processes, data literacy is pivotal to data-driven organizations so they can both take advantage of and challenge the insights an AI system can provide. If data users don’t agree or understand how to interpret their options, they might not follow the process. This can be a particularly high risk when you consider the implications this can have when it comes to cultivating a culture of data and AI ethics, and complying with data privacy standards.

Building a data-driven organization is a multifaceted undertaking spanning IT, leadership, and line of business functions. But the dividends are unmistakable. It sets the stage for enterprise-wide automation and IT. It can provide a competitive edge to organizations in their ability to quickly identify opportunities for costs savings and growth, and even unlock new business models.

Source: ibm.com

Continue reading

The importance of governance: What we’re learning from AI advances in 2022

Over the last week, millions of people around the world have interacted with OpenAI’s ChatGPT, which represents a significant advance for generative artificial intelligence (AI) and the foundation models that underpin many of these use cases. It’s a fitting way to end what has been another big year for the industry.

We’re at an exciting inflection point for AI. Adoption of AI among businesses is increasing, and research and AI development on foundation models is enabling use cases like generative AI to become even more sophisticated and powerful. The potential is vast. It can help us leverage significant amounts of data to start designing and discovering new solutions to business and societal problems such as those related to sustainability, life sciences, customer care, employee experience and many more.

These advances are simultaneously raising separate, but important discussions and questions within the industry: how can you trust the algorithms and outputs of these models? How can we ensure that these models are being used responsibly? For example, generative AI models can produce highly believable, well-structured responses so it can be hard to immediately pinpoint an incorrect response without the right subject matter expertise.

This is a dialogue that IBM is engaging in with our clients and partners every day. Advances across AI technology are happening quickly. At the same time, governments around the world are continuously evaluating and implementing new AI guidelines and AI regulation frameworks. We think that businesses have an opportunity to act now to put guardrails in place internally to govern how AI is developed and deployed.

To scale the use of responsible AI requires AI governance, the process of defining policies and establishing accountability throughout the AI lifecycle. This can also help your models adhere to principles of fairness, explainability, robustness, transparency and privacy. A comprehensive AI governance strategy encompasses people, process and technology.

Organizational AI governance processes help to decide when, where and how to use AI across the business and establish policies based on corporate values, ethical principles, regulations and laws. At IBM, we have an AI Ethics Board that supports a centralized governance, review, and decision-making process for IBM ethics policies, practices, communications, research, products and services.

AI governance technology can help implement guardrails at each stage of the AI/ML lifecycle. This includes data collection, instrumenting processes and transparent reporting to make needed information available for stakeholders. We recently launched IBM AI Governance, a solution designed to help companies get a better understanding of what’s going on below the surface of these systems. IBM AI Governance is designed to help businesses develop a consistent transparent model management process, capturing model development time, metadata, post-deployment model monitoring and customized workflows. IBM has also developed and open sourced a set of Trusted AI toolkits, including AI Fairness 360, Adversarial Robustness 360, AI Explainability 360, Uncertainty Quantification 360 and AI FactSheets 360.

In addition to discussions on the importance of AI governance, many of the advances across the industry reinforce IBM’s focus on the unique needs of AI for business. Our clients want AI that is designed for and managed by their subject matter experts, and that can be easily customized based on their domain and business priorities; that is robust with high accuracy and reliability; that operates in and navigates through siloed data in complex formats; and that is guided by principles of trust and transparency. This focus on AI for business is what guides the development of our AI software like IBM Watson Assistant and IBM Watson Discovery.

2022 has been another big year for AI with increasing adoption across the industry as well as promising new advancements. We believe that businesses that embrace AI governance early will be better positioned to responsibly harness this technology now and in the future.

Source: ibm.com

Continue reading

Maximize your data dividends with active metadata

Metadata management performs a critical role within the modern data management stack. It helps blur data silos, and empowers data and analytics teams to better understand the context and quality of data. This, in turn, builds trust in data and the decision-making to follow. However, as data volumes continue to grow, manual approaches to metadata management are sub-optimal and can result in missed opportunities. Suppose that a new data asset becomes available but remains hidden from your data consumers because of improper or inadequate tagging. How do you keep pace with growing data volumes and increased demand from data consumers and deliver real-time data governance for trusted outcomes?

It is imperative to evolve metadata management approaches to keep pace with the proliferation of enterprise data. This puts into perspective the role of active metadata management. According to Gartner, active metadata management includes a set of capabilities that enable continuous access and processing of metadata.

What is Active Metadata management?

Active metadata management uses Machine Learning to automate metadata processing and use the outcomes of that metadata analysis to help drive decisions through recommendations, alerts and more. In short, active metadata management makes data more actionable in real-time. It includes a set of capabilities that facilitate automated data discovery, improve confidence in data, and enable data protection and data governance at scale.

Common use cases for active metadata management

Improve data discovery

Research shows that up to 68% of data is not analyzed in most organizations. Knowing what data assets are available across the enterprise is key to improving data utilization. You can enable advanced data discovery with AI-driven recommendation engines that analyze active metadata and recommend new assets to data consumers based on their usage patterns.

Provide early indicators of data quality

Poor data quality is a barrier faced by organizations aspiring to be data-driven. Most data quality management approaches are reactive, triggered only when consumers complain to data teams about the integrity of datasets. Active metadata management can help with proactive data quality management. Data observability capabilities help augment trustworthy data and detect anomalies in data pipelines, allowing IT teams to quickly surface and resolve issues before they impact the business.

Regulatory and compliance

The risks of non-compliance – legal penalties, loss of reputation and customer trust – are too big to be ignored. According to the Gartner Hype Cycle for Data Privacy 2021, more than 80% of companies worldwide will face at least one privacy-focused data protection regulation by 2023. Rather than responding to each challenge individually, a proactive approach to data privacy, protection and risk management is an opportunity for organizations to build customer trust. With active metadata management, organizations can enforce data policies automatically and implement data protection rules at scale for better compliance with new data regulations.

3 benefits of an active metadata management solution

A data fabric solution connects the right data, at the right time, to the right people, from anywhere it’s needed. One of the key aspects of the IBM data fabric solution is the active metadata capabilities delivered by IBM Watson Knowledge Catalog for Cloud Pak for Data. This data catalog empowers data producers and consumers to understand, trust and protect data, and to use it confidently throughout its lifecycle.

Know your data

Ensuring that data is enriched with all the relevant context is critical for advanced data discovery and improved trust in data. Watson Knowledge Catalog helps data consumers find and understand data by offering a strong metadata foundation consisting of business terms, data classifications, and reference data backed by AI/ML-driven automation. With intelligent recommendations from IBM Watson and peers, users are empowered to find relevant assets from across the enterprise at scale. Furthermore, automated metadata enrichment built into Watson Knowledge Catalog uses machine learning to automatically assign business terms to data assets at scale. This helps users find data faster, decide if data is appropriate and can be trusted and how to work with data.

Trust your data

Complex data landscapes and resulting data silos place a time-consuming burden on data teams to govern data spread across distributed data environments and deliver trusted data.  To improve trust in data, Watson Knowledge Catalog performs data quality analysis to assign quality scores to data assets based on dimensions like data class and type violations, duplicate values, missing values, and suspect values. Custom data quality rules can then be defined to improve curation activities.  Furthermore, IBM’s partnership with MANTA brings automated data lineage capabilities to trace and analyze how data is moved and consumed across all your applications and data sources. This complements IBM’s acquisition of Databand.ai and its data observability solutions to facilitate trustworthy data by actively using historical trends and statistics to detect data anomalies in data pipelines so that IT teams can quickly surface issues before they impact the business.

Protect your data

IBM supports advanced data privacy management capabilities for dynamic enforcement of your data protection policies globally. Create data protection rules to help control access to data assets no matter where they reside, mask data at the column level and filter data rows based on row attributes. IBM can help protect sensitive and critical data through de-identification of personal information and confidential information.

Want to try out the active metadata features that allow IBM to deliver integrated quality and governance capabilities? Check out the free trial.

Access the report to read why IBM is recognized as a Leader in the 2022 Gartner® Magic Quadrant™ for Data Quality Solutions.

Source: ibm.com

Continue reading

Bridge the data literacy skills gap with data storytelling

Being a data-driven organization goes well beyond building a modern data architecture. With vast amounts of data flowing through the enterprise, the challenge lies in making sense of all of that complex information so that everyone, not just the data scientists or machine learning engineers, can interpret it for better decision-making.

For CDOs and other leaders within different lines of business, this means fostering a culture that prioritizes data literacy: the ability to read, understand, create and communicate data. Too often, data is presented in mysterious figures that are difficult for key stakeholders to understand. In fact, poor data literacy is the second-biggest internal roadblock to the success of the CDO’s office, according to the 2021 Gartner Annual Chief Data Officer survey.

To get data literacy right, organizations have to make data more approachable for non-technical experts. This means moving away from confusing charts, dashboards, graphs and complicated visuals. Instead, organizations need to humanize data and AI by creating visually compelling stories that resonate with people and transform data into actionable knowledge that drives business results.

Bridging the data literacy gap with a culture of data storytelling

Enter data storytelling, the ability to convey data not just with numbers but with engaging narratives and visuals. Creating a narrative context is important, because it brings data to life and ensures that the message it’s delivering is meaningful and relevant. Adding data visualization elements enhances the story and makes large amounts of data more digestible. When done right, data storytelling can be a powerful tool to communicate and demystify the data science.

While organizations most often relay on a combination of UX/UI designers and BI specials, when non-technical business stakeholders start developing data storytelling skills it can spark a chain reaction across teams, LOBs and the organization as a whole. In that moment, one person with data storytelling skills will lead a group of individuals to make a better, data-driven decision, But those data savvy individuals also can impart their know-how to other coworkers and inspire their teammates to hone their data storytelling skills and shape their own daily workflows.

This helps to cultivate a collaborative data-driven culture from within that gives business teams access to the strengths and skills of everyone to solve problems better and innovate faster.

The power of established data-literacy initiatives and data storytelling programs

Too often, business stakeholders blindly follow data created by algorithms. To make sure this doesn’t happen, organizations need experts who can challenge those algorithms by asking critical questions of the data and interpreting it correctly. Understanding and telling stories with data is a pivotal part of ensuring employees are still critically thinking about data, questioning it and interpreting it correctly, because it includes more people in the conversation.

The resulting unified data-driven culture brings together data visualization specialists, data scientists and software developers, executive management and other stakeholders with the goal of everyone speaking a common language made of data.

To create this culture of data literacy, organizations can start by developing a business strategy at the executive stakeholder level. Once the business strategy is clear, data leaders like the CDO can craft a data strategy that helps achieve those business goals that includes data literacy initiatives to ensure adoption and success.

Launching data-literacy initiatives and data storytelling programs will help everyone, from the C-suite to all other key stakeholders, gain the skills needed to discover data insights, trends and patterns relevant to solving business problems. Training also empowers teams to use data as a competitive differentiator.

Source: ibm.com

Continue reading

Approaches to long-term planning with IBM Planning Analytics

“If you fail to plan, you are planning to fail.” – Ben Franklin

In our collective rush to react to ever-changing marketplace dynamics and shifts in the economy, it’s easy to focus on short-term plans, to the neglect of long-term planning. Today’s leaders need to have several plans – short-term, medium-term, and long-term.

Different plans for different needs

How do these plans differ? A short-term plan is designed to show granular details for a limited time frame. This is often updated monthly, although we have some clients updating their plans on a weekly basis. One of our clients follows a process where local managers update their plans on Mondays and Tuesdays, have the regional managers review the data on Thursdays, and allow senior management to analyze and assess the data on Fridays. Each Monday they start the process over.

Most organizations utilize a medium-term plan that looks out anywhere from a few quarters to a full year. Most people will think of this as a standard monthly forecast with data at a bit more of a higher level, but still somewhat details.

A long-term plan often goes out multiple years. Many companies create a 5-year plan, although some industries such as entertainment and pharmaceutical often create 20-25 year plans. A long-term plan is a high-level view of the business. It’s not nearly as granular as short, or even medium-term plans. The plan does not get down to the level of looking at a GL account or a customer. It’s a measuring tool and a defined way of reviewing the progress of the company. In short, long-term planning helps to set the company’s direction.

The essentials of long-term planning

The long-term plan gives you guidance on how to answer several questions, including:

◉ How can we expand the company?

◉ How can we look into acquisitions?

◉ What products, geographies, and verticals can we or should we add?

◉ What products no longer make sense?

◉ How do debt payments impact cash flow?

◉ What type of labor, buildings, locations, and equipment do we need?

A long-term plan can be considered a proactive approach to risk mitigation, enabling companies to plan, think ahead, prepare for, and lessen the impact of potential negative effects. At Revelwood, we recommend two approaches to long-term planning: the growth percent approach and a driver-based approach.

We often see both of these methods used when performing long-term planning in IBM Planning Analytics with Watson:

Growth percent approach

The growth percent approach allows you to adjust groups of data (accounts, departments, etc.) by increasing or decreasing the values from the previous year. Some clients prefer to simply use a single percentage (example: reduce all expenses by 2% each year for the next five years) whereas some clients prefer to include more variation (example: reduce utilities expenses by 2% next year, by 3% the following year, and by 4% for the next three years). But no matter what level of detail is used, Planning Analytics’ powerful scripting tool will perform the entire long term plan in a matter of seconds.

Driver-based approach

A driver-based approach uses operational activity to calculate key variable revenues and expenses. This approach allows you to simplify the input by defining a set of drivers and creating calculations that use the drivers.  For example, a single driver of “units sold” can be used to immediately calculate revenue, COGS, and some of your variable expenses using the tool’s efficient calculation engine.

Mitigate risk with long-term planning

Long-term planning is your company’s assurance against planning to fail. There’s a reason why Franklin’s quote has lasted through the years. And it should be the motto of every planning team.

Source: ibm.com

Continue reading

Creating a holistic 360-degree “citizen” view with data and AI

Achieving health equity is perhaps the greatest challenge facing US public health officials today. In a 2021 report released by the Commonwealth Fund, the nation ranked last among high-income countries in access to healthcare and equity, despite spending a far greater share of its GDP on healthcare.

Healthcare disparities are closely linked to race, ethnicity, gender and other demographic and socioeconomic issues surrounding access, cost and quality of care. Health inequities in the US came into sharp relief during the COVID-19 pandemic: Analyses of federal, state and local healthcare data show that people of color experienced a disproportionate burden of cases and deaths.

But there is promising news. The recent crisis not only highlighted the critical need to focus more on health equity but also revealed how tapping into data-driven technologies can better ensure equity for marginalized groups.

In 2020, IBM collaborated with the Rhode Island Department of Health, uncovering existing and emerging data patterns to aid the agency’s overall response to the health crisis. This work resulted in real-time, data-driven decisions that identified pandemic-fueled disparities such as lack of access to vaccines. Ultimately, the project led to more equitable emergency response services in the Rhode Island regions that needed it most.

Today state health departments around the country are taking the data-leveraging lessons learned during the pandemic and applying them to an array of public health crises affecting underserved groups. Health departments are focusing on issues such as food insecurity, unwanted pregnancies, increased suicide rates and opioid addiction. Thanks to innovations in data analytics and AI, leaders can make smarter, faster and more efficient decisions to improve public health outcomes and advance health equity.

Creating a citizen 360 view

The journey begins with building a data fabric architecture to ensure quality data can be accessed by the right people at the right time, no matter where it resides. The key is making sure all this data is transparent and responsibly governed for privacy and security.

A data fabric facilitates the end-to-end integration of various data pipelines and cloud environments by using intelligent and automated systems. It also provides a strong foundation for 360-degree views of customers, or in this case citizens rather than customers.

In B2B or B2C circles, a 360-degree view of customers or citizens offers a holistic, comprehensive picture of a person based on data collected from all touch points. This drives business value by creating more effective outcomes as well as more personalized customer experiences. For instance, this data infrastructure enables a state health workforce to better understand the overall healthcare landscape and subsequently improve individual care and address inequities.

Achieving data literacy with storytelling and visualization

Collecting massive amounts of data presents a common issue for both private and public enterprises: how to make sense of all that data.

Part of data storytelling involves data visualization, the process of analyzing large amounts of data and communicating the results in a visual context. But strong storytelling must go beyond presenting data in the form of charts, graphs and tables.

For instance, state health departments comprising many stakeholders and players need to create a compelling storyline and consistent messaging around their data, so they can communicate it effectively to their entire workforce.

Keeping the citizen front and center

Data and trustworthy AI also provide predictive analytics for insights that can solve some of the most pressing health issues, including hunger and food insecurity.

For example, the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), a federal food assistance program that operates through state health departments and local agencies, has seen decreasing enrollment over the past decade despite a sharp rebound in poverty levels. Suspected factors include slow modernization — until recently, all WIC benefits were still delivered as paper vouchers — and persistent stigma against federal assistance. Providing assistance depends on identifying and addressing these and other factors.

The WIC Enrollment Collaboration Act of 2020 calls for state health departments to count unenrolled WIC-eligible families. A data fabric with a 360-degreee view can help that count. It can also help states build and deploy referral mechanisms and conduct a comprehensive outreach campaign (also detailed in the Act). Working together, states can use data to assess and improve access to WIC and better limit food hardship.

Throughout the US, state departments of health, education and behavioral health are using data to overcome other health crises, including the opioid and suicide epidemics. A centralized data hub provides a powerful public health crisis response system that allows for collaboration across government branches and state lines. Such multi-pronged efforts are closing the gap in critical information, shedding light on how and why disparities occur and paving the way to better health equity for all.

Today IBM is working with state health departments to accelerate their digital transformations in the areas of overall governance, operations, automation, data insights and more.

Source: ibm.com

Continue reading

IBM named a leader in the 2022 Gartner® Magic Quadrant™ for Data Quality Solutions

Data is the new oil and organizations of all stripes are tapping this resource to fuel growth. However, data quality and consistency are one of the top barriers faced by organizations in their quest to become more data-driven. So, it is imperative to have a clear data quality strategy that relies on proactive data quality management as data moves from producers to consumers.

Unlock quality data with IBM

We are excited to share that Gartner recently named IBM a Leader in the 2022 Gartner® Magic Quadrant™ for Data Quality Solutions.

Access the full report here.

We believe, this is a testament to IBM’s vision to empower data professionals with trusted information through data quality capabilities including data cleansing, data lineage, data observability, and master data management.

IBM recently expanded its data quality capabilities with the acquisition of Databand.ai and its leading data observability offerings. This complements IBM’s partnership with MANTA to integrate automated data lineage capabilities from MANTA with IBM Watson Knowledge Catalog on Cloud Pak for Data.

Why does data quality matter across the data lifecycle?

Data quality issues can have far-reaching consequences across the lifecycle of data:

1. Analytics and AI

When a sophisticated AI/ML model confronts bad-quality data, it is the latter that usually wins. As organizations increasingly rely on AI/ML for critical business decisions, the role of a trusted data foundation that delivers high-quality data is paramount. So, it is important to provide data consumers with a curated set of high-quality data and allow them to search for relevant data through a well-defined data catalog.

2. Data Engineering

A research survey points out that data engineers spend two days per week firefighting bad data. This could be because a lot of the current data quality approaches are reactive, triggered only when data consumers complain about data quality. Once poor-quality data moves from data sources into downstream processes, it gets challenging to remediate quality issues. A smarter approach would be to plug data quality issues upstream through active monitoring and automated data cleansing at the source. Data observability capability makes data quality checks upstream possible.

3. Data Governance

Ensuring data quality is critical for data governance initiatives. Increasingly enterprise data is spread across multiple environments which contributes to inconsistent data silos that complicate data governance initiatives and create data integrity issues that could impact Business Intelligence and analytics applications. It is critical to promote a common business language across the enterprise to break down these silos. One effective way to identify bad-quality data before it flows into downstream processes is with the use of active metadata to foster greater understanding and trust in data and ensure that only high-quality data makes its way to data consumers. Equally important is the ability to understand data lineage by tracking the flow of data back to its source which can prove handy when remediating data quality issues.

IBM’s holistic approach to Data Quality

With a strong end-to-end data management experience combined with innovation in metadata and AI-driven automation, IBM differentiates itself by offering integrated quality and governance capabilities.

IBM Watson Knowledge Catalog, QualityStage, and Match360 services on Cloud Pak for Data offer a composable data quality solution with an easy way to start small and expand your data quality program across the full enterprise data ecosystem.  Watson Knowledge Catalog serves as an automated, metadata-driven foundation that assigns data quality scores to assets and improves curation through automated data quality rules. The solution offers out-of-the-box automation rules to simplify the addressing of data quality issues.

With the recent acquisition of Databand.ai,  a leading provider of data observability solutions, IBM can elevate traditional DataOps by using historical trends to compute statistics about data workloads and data pipelines directly at the source, determining if they are working, and pinpointing where any problems may exist. IBM’s partnership with Manta for automated data lineage capabilities further strengthens its ability to help clients find, track and prevent issues closer to the source and for a more streamlined operational approach to managing data.

IBM offers a wide range of capabilities necessary for end-to-end data quality management including data profiling (both at rest and in-flight), data cleansing, data monitoring, data matching (discovering duplicated records or linking master records), and data enrichment to ensure data consumers have access to high-quality data.

Gartner does not endorse any vendor, product or service depicted in its research publications and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner’s research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.GARTNER and Magic Quadrant are registered trademarks and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and are used herein with permission. All rights reserved.

Source: ibm.com

Continue reading