We’ve all seen the hype around Enterprise Big Data and AI build up over the last few years, culminating in a record year of investments, conferences and implementations in 2017. But how real is AI when it comes to building value for your business today and over the next five years?

Although we are certainly many years away from a human-like AI as we see in the movies; today, narrow or domain-specific AI technologies are already making an impact on bottom lines. Companies that have been smart about adoption and able to quietly implement AI-aided solutions into various functions such as Demand Planning and Inventory Management, Back Office Processes, Sales and Marketing are reaping the benefits.

Because AI can help companies find competitive advantages, demand is increasing at an incredible pace. New companies offering AI enabled software, and other technologies seem to pop up almost daily. Considering the amount of money and brainpower poured into AI research, it won’t be long until commercializing and monetizing data using AI as well as transforming internal processes becomes a necessity to remain competitive.

According to the recently published Teradata report State Of Artifical Intelligence For Enterprises, the majority  “see AI as being able to revolutionize their businesses, automating repetitive processes & tasks and delivering new strategic insights currently not available.”

But with most enterprise software initiatives taking on average 21 months to implement and with Big Data and AI being at the complex end of the spectrum, it is no surprise that 91% see barriers ahead with lack of IT infrastructure (40%) and lack of talent (34%) as the most significant.

So how do you quickly adopt AI successfully across different business functions, driving real and immediate ROI?

AI as a Service

AI Software As a Service (SaaS) adoption is a clear trend that is taking hold in enterprise technology stacks. Adopting SaaS solutions can help companies smooth out their revenues, leading to more resilient and flexible organizations, ultimately allowing a company to deliver better service and products to their clients. With a shortage of talent in this arena and the large data sets required to effectively train artificial intelligence algorithms and implement them into production software, the SaaS model has clear advantages versus trying to develop all capabilities in-house.

Definitive Advantages

The reasons for moving to SaaS offerings can be different for each organization. One of the primary drivers is the potential to create a technology advantage over established competitors and potential disruptors.  Others find they’re increasingly dissatisfied with the way their legacy functions and processes run, and want a better and faster way to see improvements.

Services are defined based on business results and can be expected to produce value quickly, be flexible, implemented quickly, and paid for based on value, business outcomes, or on a seat/consumption basis. This approach leaves more room for pivoting if the ROI is not there as promised, in contrast to traditional capital investment projects where teams often fall prey to the sunk cost fallacy or have a hard time measuring the ROI of their investment.

Enterprises that transition to this model will have a definitive advantage over those that don’t. Companies that don’t shift to aaS models will see their ability to compete diminished, and the same can be said about leveraging AI enabled technologies such as Robotic Process Automation and Automated Insights Generation to name a couple of tangible applications of AI in the enterprise today.

A SaaS tech stack also offers a company greater agility. Traditional industries are consolidating amid increasing mergers and acquisitions, and that means becoming more agile and lean to compete and continue to grow. Service-based models allow companies to trim infrastructure, creating flexibility to scale up or down depending on business needs.

A SaaS model also enables better analytics to derive business insight and help make performance improvements. With clear and contained costs and sometimes built-in analytics capabilities, it is easier than ever to evaluate business results and ROI of investments in services vs. traditional Capex expenditures.

Getting There

Determining how to start adopting AI technologies as well as transitioning to a SaaS and multi-cloud based stack is not necessarily easy. Where to start? With a single problem, department or business need, or do you embark on an enterprise-wide effort?

It can be as simple as starting small with low-hanging fruit and then expanding from there. Is there a department that is last on the priority list for IT but could make some significant gains if given the right tools today? Is there an apparent cost, margin or process that can be identified for measurable improvement? Companies that have seen immediate success often start small and then build on that success. Technology moves too fast these days to allow for extensive planning and execution timelines.

No matter how they get there, in the long run, businesses that transition to service-based models have incomes that are more consistent over time, allowing them to make better and more agile decisions that lead to robustness, flexibility and therefore long-term sustainability.

There are many trends coming to the foreground of AI, machine learning, and business intelligence. This article will be talking briefly about some of these trends and why they are coming to light. A link to the in-depth report by Tableau can be found at the bottom of the page.

Do not Fear AI

Is AI the destructive force that will destroy all jobs and the world as we know it? The media and Hollywood have depicted AI as such, however this is not the case at all. At this point in time, machine learning and AI has become a daily tool in business intelligence. These tools are giving time back to their human Analyst counterparts. Analysts are using machine learning and AI software to better understand their company’s data in a more timely fashion.

Liberal Arts Impact on AI

In the upcoming months Liberal Arts will be playing a bigger role in the building of AI and machine learning software. Data scientists are realizing they not only need the data analyzed to be accurate but also tell a story that anyone can understand, including those without a technical background.

NLP (Natural Language Processing) Promise

NLP refers to the way we interact with the AI through the UI (user interface). Companies are beginning to want all level of employees to have access to the data provided by their AI software. The problem many of these companies face is that most of their employees do not have a technical background and no idea how to query a piece of data. This is where NLP comes into play; AI software can process queries in natural language instead of using specific codes. e.g. I want to know the Sales for Item “001”  by day at Store “2045”

Multi-Cloud Capabilities

The move to multi-cloud storage is becoming an ever-increasing desire within big companies. Companies don’t want to be limited to one storage method that may not provide the best performance for their data needs. Though multi-cloud architecture has many benefits, it also has its costs, one of which being the actual overhead cost of running this type of multi-cloud environment.

Rise of the CDO (Chief Data Officer)

With understanding data and analytics becoming a core competency more and more companies are creating a position of CDO. This position allows them to join the C-suite with the CEO, CTO and CIO. This new position gives the CDO the ability to attend the C-level meetings and actually affect change within the company. Due to the creation of the CDO position, companies are showing just how important it is to understand their data and manage it successfully.

Crowdsourcing Governance

Crowdsourcing governance is a fancy term for allowing customers to shape who has access to specific data within a company using self-service analytics. It gets the right information into the right hands while keeping that same information out of the wrong hands.

Data Insurance

Data is more valuable than ever. We have seen countless data breaches over the last few years and will most likely see many more. With customer data becoming so valuable we are going to see a rise in data insurance. This insurance will protect companies from being responsible for a breach of their customer data.

Data Engineering Roles

As data analysis software continues to grow in use and value we will see a rise in data engineering roles over the next several years. Data engineers will begin to transform from more architecture-centric roles to a more user-centric approach within their organizations.

Location of Things

“Location of things” is in connection to IoT (internet of things). We are seeing companies trying to capture location-based data from IoT devices. Gartner, predicts there will be 2.4 billion IoT devices online by 2020. The problem is that companies are trying to collect and compile all this location data within their internal data structures, while most of these structures are not capable of accepting that quantity of data. This is going to lead to great innovations for IoT data storage.

Academics Investments

With data analytics growing in all industries the demand for future data scientists will continue to grow. Due to this high demand for data engineers and data scientists we will begin to see more and more universities offering some sort of academic training in these categories over the next several years.

 

Read the full report by Tableau Here:

https://www.tableau.com/reports/business-intelligence-trends#ai

We would like to thank Clariden Global for hosting the “Einstein AI, Deep Learning & SuperIntelligence Summit” and inviting Nikki to speak at the post-summit seminars about AI Assistants in Business. The conversations with attendees and other speakers were high quality and it was great fun to ponder the future of these cutting-edge technologies.

If you missed the talk and would like to see the slides get in touch!

A huge thanks to everyone that came out to our first workshop event in Bentonville, AR! We had so much fun doing a Walmart Store 1 walk-thru (did you know there is a Dunkin Donuts inside!!), visiting customers, and holding our first Analytics Edge Workshop at the 21c Hotel.

The Workshop was incredibly well attended and we had some great dialogue with suppliers and others in the community about the potential of AI for business process automation and how easy access to Analytics can give anyone in an organization the opportunity to be strategic in their roles.

We also want to give a big thanks to John Daly of Sony Pictures Home Entertainment for joining us as a guest speaker, it is always inspiring to hear him talk about the transformation they have been able to achieve in such a short time!

We look forward to being back in Bentonville very soon!

In a 2016 research reportWhy Artificial Intelligence is the Future of Growth, Accenture found that adoption of artificial intelligence tech across all industries may double economic growth rates by 2035. AI investment is expected to increase labor productivity by 40 percent. In fact, 70 percent of executives say they plan to “significantly increase” AI investment.

In the realm of inventory and supply-chain management, AI adoption, specifically the use of optimization algorithms, is revolutionizing inventory agility – reducing stock depletions and maximizing stock levels.

“The use of AI in supply chains is helping businesses innovate rapidly by reducing the time to market and evolve by establishing an agile supply chain capable of foreseeing and dealing with uncertainties,” says Accenture Managing Director Manish Chandra. “AI armed with predictive analytics can analyze massive amounts of data generated by the supply chains and help organizations move to a more proactive form of supply chain management.”

Supply chain processes generate giga-tons of data, and AI can deploy predictive analytics to make sense of it all. Freshly updated and analyzed data then builds a solid foundation when it comes to real-time vision and information flow. Every key player across the supply chain is empowered with the best data and maximizes it accordingly.

AI is no longer an “ain’t-it-cool” innovation in the industry but rather a necessity. With the erosion of the brick-and-mortar model and rise of real-time consumer expectations, supply chain/inventory management practices must embrace machine learning that far outpaces the speed of human thought and action. Consider these stats from the 2017 MHI Industry report concerning the speed of supply-chain transactions from just one e-tailer on Black Friday:

“A reported 426 orders per second were generated from the website throughout the day. That equates to over 36 million order transactions, an estimated 250 million picking lines at the distribution centers (DC), 40 million DC package loading scans, 40 million inbound sortation hub scans, 40 million outbound sortation hub scans, 40 million inbound regional sortation facility scans and 40 million outbound delivery truck scans.”

How should industry leaders respond? The answer, according to the report, is clear. Supply-chain companies must embed “analysis, data, and reasoning into the decision-making process. Position analytics as a core capability across the entire organization, from strategic planners through line workers, providing insight at the point of action.”

As Accenture economic research director Mark Purdy concludes, companies that survive will fully invest in the potential power of AI going forward: “To fulfill the promise of AI, relevant stakeholders must be thoroughly prepared – intellectually, technologically, politically, ethically and socially – to address the benefits and challenges that can arise as artificial intelligence becomes more integrated in our daily lives.”

Analogue: Scalability in Data Usage

At the intersection of big data and machine learning are patterns and analyses that reveal trends and causes. To use healthcare as an example, sensors built into wearable medical devices open windows to improved, individualized healthcare based on a rapidly expanding set of clinical, lab, physiological, and personal data. (A patient diagnosed with hypertension might wear a device that sends information to an application that detects ongoing changes in blood pressure, respiration, or other conditions in real time and alerts a physician when anomalies occur.)

Predictive data technology moves past the goal of gaining insights and into the realm of insights on insights: namely, choosing the trends that require action. If the information received from the wearable monitor is utilized as cross-channel data, the challenge becomes making sense of the insight gained from the data and selecting the appropriate action. With this, the perspective may move from a simple focus on the instantaneous symptoms and treatment of hypertension to a holistic view of the patient’s respiratory, renal, and other systems’ response to standardized treatment.

The Human Factor

The relevance of obtaining cross-channel data from a hypertension patient is most apparent in the universal desire for individualized care. Scalable machine learning searches for efficient algorithms that can work with any amount of data and detect hidden insights. These insights yield logical, adaptive reasoning in performing specific actions, without consuming greater amounts of computing resources. Limits do exist, but predictive data technology adds another dimension to the interpretation of vast data sets. One that, in a business context, means greater efficiency and more thorough self-evaluation on a global scale.0

In the marketplace, insights gained from cross-channel data emphasize the individual’s ability to change. While individuals may defy—with varying levels of deliberateness—predictability, machine learning and predictive data technology take an unrestrained, multi-dimensional view of preferences, real-time behavioral patterns, and possible intent. Thus the view of the “customer journey” is expanded: and a mass of stops at a big box store from which a correlation would have normally been determined in retrospect is now a targeted real-time marketing effort—with the intuition to make progressively better use of progressively expanding data.0

Moving to a New Meaning

Terms like “segmentation analysis” and “adaptive marketing” are themselves harbingers of a system that will soon replace the marketing philosophies of old. However, these new practices may themselves prove to be stepping stones to an even broader view of personalized marketing. Real freedom from scale is measured over time: through predictive data technology that offers personalized strategies for small businesses, large businesses, and corporations as they grow. This new outlook recognizes the consumer’s awareness of the marketplace and the complexity of their decisions, providing insights into profit margins based not only on the instantaneous relationship between product and cost, but also by an adaptive view of long-term customer behavior and loyalty.

There are many trends coming to the foreground of AI, machine learning, and business intelligence in 2017. This article will be talking briefly about these trends and a link to the in-depth report by Tableau can be found at the bottom of the page.

BI (Business Intelligence) the New Norm

In 2017 we will see a trend of more and more companies using modern business intelligence, allowing analytics to be performed by all employees, not just data scientists and engineers.

Collaboration between Machines and Humans Strengthens

The collaboration and sharing of data is going to move from one-direction, spreadsheets and emails, to an interactive flow of data between multiple parties and their live data stream.

Data Will Become Equal

All data will be equally accessible and understandable. We will be able to access all our data without the worry of it being stored in the same format.

Anyone will be able to Data Prep.

Just as self-service analytics is becoming more accessible to non-technical employees, so will the ability to understand and prep data without the need of a technical background.

Imbedded BI is Allowing Analytics to Grow Everywhere

Business applications like Salesforce are placing analytic tools in the hands of people never before exposed to data. These tools are extending the reach of analytics in our day-to-day lives and we most likely are unaware that we are using them.

Work with Data in a Natural Way

In the next year we will see people being able to access and communicate with their data in a more natural way. We will see this more with the integration of natural language interfaces within AI networks.

Cloud Based Analytics

With data being stored in the cloud we will soon see analytics being conducted there as well. Cloud analytics will be faster and able to scale at a much quicker pace.

Data Literacy will Become a Necessity 

With Data analytics and predictive analysis moving to the mainstream we will see a need for all level of employees needing to be able to read and understand their company’s data.

 

Read the full report by Tableau Here:

https://www.tableau.com/learn/whitepapers/top-10-business-intelligence-trends-2017