Artificial Intelligence

Introducing Aera’s Cognitive Technology

Enabling The Self-Driving Enterprise

Cognitive capabilities are highly valued in human beings. They make people smart, and smart is good. According to the Oxford Dictionary, cognition is “the mental action or process of acquiring knowledge and understanding through thought, experience, and the senses.” Yet as automation becomes more and more prevalent, we expect more and more functions and processes to be performed without human assistance. Can technology really imitate human cognition? Why not? After all, we live in a world where self-driving cars, although not yet ubiquitous, are a reality. And in a world where terabytes of data are being replaced with zettabytes, is it even possible for a human to process data at the speed and granularity necessary for timely, data-driven decisions?

Enterprise applications have been used to streamline and automate transactional processes for several decades now, particularly where simple and straight forward rules can be applied. When inventory falls below safety stock, order more. But how do you know when to change safety stock? How do you balance inventory across your distribution network or work off excess inventory? How accurate is your forecast? Is it possible to automate the cognitive functions that understand (recognize patterns and learn from the past), predict the future, and not only make recommendations, but also take action? Aera Technology not only thinks it is possible, it is delivering on that promise today to enable the Self-Driving Enterprise.

Aera is quite a unique kind of company. Headquartered in Mountain View, California, it serves some of the world’s largest enterprises from its global offices located in San Francisco, Portland, Bucharest, Cluj-Napoca, Paris, Munich, London, and Pune. Using proprietary data crawling, industry models, machine learning and artificial intelligence, Aera’s goal is to revolutionize how people relate to data and how organizations function. It offers what it calls a “cognitive operating system.”

The Self-Driving Enterprise

Aera starts with the premise that if built-in intelligence can drive a car, then it should be able to drive a company. Like a self-driving car, a self-driving enterprise must connect all the different data points both inside (engine, accelerator, steering wheel, brakes) and outside (roadways and road conditions, other vehicles, pedestrians). It must do all this in real-time, because speed and direction changes must occur immediately as any of those conditions change. And it must be always on and always thinking. No snoozing at the wheel allowed. It also must be able to operate autonomously. With no driver, a self-driving car has to take action without being told what to do.

A self-driving enterprise will still have humans at the helm. Aera is not setting out to eliminate the decision-makers, but it is trying to make them smarter and more effective, able to use all the data available, not just the usual subset contained in an enterprise resource planning (ERP) solution.

If this has you curious to learn more, click here to read the full report.

Tagged , , , , , , , , , , ,

Workday: Getting Smarter and Smarter

Enter the Age of Intelligence

In a recent Mint Jutras report, “How Smart Are Your Enterprise Applications?” we outlined some of the different ways solution providers are adding a new level of intelligence to their offerings… or not. While “intelligence” has become the holy grail of enterprise applications of late, not all vendors are delivering on the promise of smarter applications. For some, it’s just the latest buzzword added to their marketing collateral and some are simply playing catch up to current next generation applications. Others are taking their first baby steps, but a select few are truly entering the “age of intelligence.”

Where is Workday along this progression? Since its inception in 2005, it has never been a company that over-inflated its capabilities with bravado and marketing spin. Born in the cloud and built on a next-generation platform that continues to evolve, Workday also never had to play catch-up. And the first steps it took in moving into the age of intelligence were not baby steps, but instead bold ones, including some strategic acquisitions.

Workday’s acquisition of Identified in 2014 was an important step in incorporating predictive analytics and machine learning into its portfolio. In 2015 it acquired Gridcraft and last year it acquired Platfora. With both of these acquisitions, Workday sought to build insights [read intelligence] directly into its applications. More recently its benchmarking capabilities take insight and intelligence to another whole level by putting Data as a Service (DaaS) in the context of your business performance, in comparison to your peers. And Workday has opened the doors to more innovation from a broader community by making its Workday Cloud Platform available beyond its own development team.

It is clear Workday is getting smarter and smarter with each new release.

Smart, Smarter, Smartest

So, what does it take to make an enterprise application smart? In our previous report we distinguished different levels of intelligence:

  • Smart: We concluded any enterprise application is smart in that it’s not dumb. It can follow instructions – instructions like, IF <this condition> THEN <do this> ELSE <do that>. Business applications have been built on IF THEN ELSE statements since the earliest computer programs were developed. Workday applications are no exception and indeed, they can now go beyond simply following specific instructions. They are starting to learn to take some simple rule-based actions on their own. For example, the recruiting module is smart enough to decline any outstanding applicants once a position is filled, and yet keep them on file to review when other vacancies open up.
  • Smarter: To make an application smarter, you need to make it easier to use and better at communicating. Progressive releases of Workday have made the user experience very compelling while also adding more and more insights. Workday has also borrowed concepts from consumer technology, putting more power in the hands of users using mobile devices, not only alerting managers to exceptions, issues and required approvals, but allowing them to take immediate action. Workday Talk provides a “chat” capability modeled after social media. Participants can follow conversations attached to business objects like sales orders, customers or products. Groups and teams can be assembled to foster collaboration. When people are better informed, they can make more intelligent decisions, faster.
  • Smartest: But the smartest applications today combine the pattern recognition capabilities of machine learning to produce artificial intelligence (AI) and predict the future. The highest level of intelligence will be achieved in combining a variety of technologies together: AI, deep machine learning, Natural Language Processing (NLP), image recognition and predictive analytics are all at the forefront of this movement. And Workday has all these technologies in its kit bag. It has already taken some initial steps in leveraging them. For example, it has embedded machine learning capabilities into its Talent Insights to identify retention risk. Look for more use cases to be delivered using data from both inside and outside of Workday.

It is quite clear that Workday’s Human Capital Management (HCM), Financial Management, Student Management and Planning solutions are smarter than your average enterprise applications. Let’s dig a little deeper into some ways they will get even smarter.

Building Insights In: Prism Analytics

Good reporting is a necessary backbone of applications like HCM and financial management. Reports provide a historical perspective, help you assess your current position and answer questions you have about your performance. But analytics provide a deeper level of understanding and help you ask the right questions. Analytics are iterative by nature. You start with a question, issue or problem: Sales are down. Reports might tell you what regions or products are problematic, but you won’t really know why until you drill down, and you are never quite sure what path you need to take until you find out more. And you won’t even be prompted to investigate until you already have a problem.

Predictive analytics help you anticipate conditions, prompting you to investigate a situation before the problem rears its head. You would like to be able to conduct this kind of investigative work right in the familiar environment of the solution running your business. But it is even more powerful when you can look beyond the structured data that resides within your enterprise applications. Workday has woven the technology acquired from Platfora, into the fabric of its solution, rather than bolting on components. And yet Workday Prism Analytics will not be limited to Workday data, but will also bring in non-Workday data, which can then be presented through Workday reports, scorecards, and dashboards for analysis.

Typically this type of mix of data requires data preparation to be done by a data administrator with the technical skills needed to load the external data, cleanse and prepare it and then create reports, queries and/or dashboards. This activity doesn’t go away with Workday Prism Analytics, but it is simplified enough for a technical business user to perform – and perform quickly enough to be of value. And the data can be blended with, transformed and enriched by your transactional system of record (Workday data). In doing so Workday has struck a nice balance between having a super powerful tool on the back end but also super easy to use on the front end, avoiding the usual trade-offs.

Workday is in the early stages of delivering this, and also has plans down the road for data discovery. Data discovery typically goes after big data in search of patterns that may not be intuitively obvious. Using the right visualization tools, it helps you understand which data is most relevant to your problem, even if you don’t know exactly what to ask for.

Benchmarking Performance with Data as a Service (DaaS)

It takes a different kind of intelligence gathering to understand your business performance in relation to others in similar roles or industries. As a multi-tenant SaaS solution provider, Workday is in a unique position to provide you with access to this kind of comparative data. But of course, you must be willing to give, in order to receive. Workday needs permission to use this data, but paraphrasing the words of Workday leadership: We don’t take customers’ data. They give it to us.

Workday sits on a large volume of data collected from hundreds of customers subscribing to its software. This is data that can be invaluable to the entire Workday community for benchmarking against peers. Customers must opt in to contribute secured aggregated data. In turn, they receive benchmarks. Today this Data as a Service (DaaS) is available for customers to explore Workday usage and HCM results, including workforce composition, diversity, turnover, etc. Financial management data is coming soon. Within the first three weeks of this service being available, Workday reported 100 customers had opted in and contributed data. Obviously, as this number grows, so will the value of the data.

Expect more from Workday along these lines in the future, including data from other sources (private and public) not included in Workday.

Machine Learning and AI

Of course the availability of a growing volume and diversity of data opens the door for machine learning and therefore artificial intelligence. Workday’s acquisition of Identified in 2014 was an important step in incorporating predictive analytics and machine learning into its repertoire of capabilities. Identified’s patented SYMAN (Systematic Mass Normalization) technology mines Facebook for social data and then uses artificial intelligence to transform that data into professional intelligence. The “learning” comes from continued use, validating predictions with outcomes from Workday employee data on performance and retention.

Workday released Workday Talent Insights in 2015, identifying retention risk and delivering a talent scorecard. Through this introduction Workday learned that customers prefer an embedded experience, not a standalone application and that the overall user experience is paramount, along with access to data for training algorithms.

The Power of a Platform

Since it was founded in 2005, Workday has always insisted it was (and is) an applications company, rather than a technology company. It has always offered cloud-based business solutions. While it built these applications on a solid and modern platform, it always resisted the urging of pundits and industry observers to become a “platform” company. Until now.

The Next Chapter for Workday

Now it will be both a “platform” player as well as a business solution provider. The Workday Cloud Platform was soft launched a few months ago with selected service partners. Built on the principles of openness, Workday will provide the tools needed to manage the complete application life cycle, with data modeling and a single Application Programming Interface (API) point of integration.

So how does this make Workday applications smarter? Of course there are no guarantees, but by opening up the platform, along with all the presentation services, conversation services, and analytics Workday uses to make its solutions smarter, the level of intelligence is more likely to deepen. The Platform will include both Workday Talk (NLP) and BOT for anomaly detection.

So, what are developers building on the platform? Here are a few examples:

  • Talent Mobility, allowing employees to visualize career opportunities and connect with employees across globe.
  • ID Services to manage security badges
  • Supplier requisitioning that allows suppliers to directly populate data in Workday
  • Safety services management


The Innovation Keynote at the 2017 Workday Rising Event was entitled “The Age of Intelligence.” The Keynote was presented by Mike McNamara, the CEO of one of Workday’s largest customers, Flex (a contract manufacturer formerly known as Flextronics). In his opening remarks, Mr. McNamara summed up this new age by saying, “Today it’s not about controlling land and resources, but rather about applying intelligence.”

In many ways, intelligence is a new currency in the global, digital economy. And yet, when most solution providers today talk about intelligent applications, they often simply mean new ways of interacting with the solution and analytics that help you derive more and better insights from the data. But this is the minimum you should expect today. Workday has aggressively taken steps towards real intelligence, through acquisition and its own development efforts. Workday Prism Analytics, Benchmarking and DaaS, machine learning, natural language processing and the Workday Cloud Platform all combine to provide powerful insights and intelligence, not through separate bolt-on tools, but embedded in a single solution.

If your current solutions are not headed down the path towards intelligent applications, if you are starting to look for new, smarter ones, Workday is a good place to start.

Tagged , , , , , , , , , , , , , ,