Artificial Intelligence in Enterprise – Robo-Management Consulting is a Reality

Artificial Intelligence in Enterprise – Robo-Management Consulting is a Reality

Introduction

Much of the recent AI revolution has been focused on automation through big data and/or sensors and feedback into neural networks. The resulting applications are highly valuable to businesses and consumers. They improve quality of life by optimizing labor and resources. However, these applications fall short when it comes to handling human reasoning. Much of the rationale behind the operation of these systems are implicitly embedded in the data. In this article, I explored a different approach to AI where a machine uses reasoning to enable humans to solve new problems and understand new subjects. I have implemented such a system that – to a high degree – can automate the role of a management consultant. Using principles of Meta-Vision and Bionic Fusion, an AI system can automate much of the mentally-intensive work a management consultant performs. This automation results in a faster pathway to insight and elevates tactical strategic planning & execution. I call this Bionic Fusion: “Robo-Management Consulting.”

Machine learning, quantitative data, and the importance of reasoning

Data science enables us to learn by observing data behavior. Machine learning is a popular technique used to discover how common data patterns relate to common outcomes. This enables data scientists to predict outcomes with present data – a strategic benefit driving the adoption of machine learning across many enterprises. This approach, however, has limitations; machine learning can only learn from the past. Humans are dynamic; change is constant. Machine learning models trained and tested for accuracy against historical data don’t know what to do when faced with zero-day scenarios – new variables, unknown outcomes. When an application works with human behavior, the AI must account for human rationale. If presented with a conference dialogue or transcript, the application must understand context, sentiment, and rationale based on the situation.

In recent years, many analytical solutions and tools have focused on quantitative data to navigate and extract insights. Quantitative data, however, is only a measure of human behavior, not rationale. For example, ‘Same Store Sales’ is a metric often used in the retail industry. Machine learning models may recognize a decline, but will miss the underlying reasons driving that change – critical insights for executives managing a turnaround or competitors looking for a weakness to exploit. Identifying and understanding the root-cause is critical to successful business execution. The value of rationale analysis is just as important – if not more – than quantitative analysis in the formulation of tactical strategy.

An Implementation Using Reasoning Models

In our implementation of our rationale analytics, I look at the science of rationale as a determining factor in selecting algorithms for analysis.

Depending on the nature of the problem, I use three different reasoning models for rationale analysis:

  1.     If a given premise is known, I use a deductive reasoning model. Known premises are inferences drawn within the scope of propositional logic.

  2.     If a given premise is unknown, I create a hypothesis. I then use inductive reasoning. In this case, the causation model is a new hypothesis. I will not apply the newly learned premise in deductive reasoning until it is accumulated into a class of common truth.

  3.     When the observation is incomplete, I need to hypothesize the missing piece(s) of the puzzle with an educated guess. For this type of situation, I use abductive reasoning. I will then use our data lake to drawn reference and validate our causation model to complete the rationale.

In my previous blogs, I have discussed some of the novel technologies that I have developed for performing these tasks. I am unaware of any open-source implementation of these principles. For the purpose of discussion, I use our SaaS analytics service to develop Robo-Management Consulting and create management analytics reports with the help of artificial intelligence. Using “Context Discriminant”, I am able to extract important subjects and supporting facts from a corpus to get a high-level view with “Meta-Vision”. The “Meta-Vision” graphically shows us the attributes of relationships between the “Machine Generated Hashtags” and supporting facts in original context.

Through this process, our supporting fact model is transformed into a propositional causation model that corroborates the premises using business intelligence from our data lake. By combining the rationales of both the original corpus and the corresponding corpus from BI, I created a rationale model. The resulting Meta-Vision is then used to obtain insights and solutions to complex problems.


Artificial Intelligence in Enterprise – Robo-Management Consulting is a Reality

Artificial Intelligence in Enterprise – Meta-Vision improves outlook and quarterly earnings call for publicly traded companies

Artificial Intelligence in Enterprise – Meta-Vision improves outlook and quarterly earnings call for publicly traded companies

Introduction

The quarterly earnings call is a critical event for publicly traded companies. Each call serves multiple purposes. It is both an important source of information for investors and an opportunity for a company to present a narrative of operational performance, financial health, and strategic vision in their own terms. It’s also an ideal opportunity for executives seeking to manage and optimize outcomes.

 

The advent of AI makes it plausible for the Chief Financial Officer (CFO) or Chief Executive Officer (CEO) to predict the outcome of a conference call by means of business intelligence (BI) analysis. Working with unstructured textual data, an AI engine can discover, investigate, and draw relationships between context to provide a meta-view of a company’s operations, performance, strategy, and alignment with wider industry trends. This enables a company to maximize the outcome of a quarterly earnings call presentation.

 

The Earnings Call Challenge of Publicly Traded Companies

Most enterprise implementations of AI are predominantly based on machine learning (ML) – solving problems using past experience, known variables and outcomes. This type of AI tends to make humans more passive when a complete solution to a problem is formulated and rendered by machine. We see examples in applications such as Google Lens, Alexa/Echo, Apple/Siri. These applications use automation to perform jobs humans can do. The only difference is in cost, time, and precision.

 

When applied to a quarterly earnings presentation or company outlook, a ML approach would be hindered by the lack of relevant historical data. Moreover, while knowledge base models can play an important role in many AI applications, these models generally fail when faced with unprecedented scenarios and “one-off” variables. For example, IBM has executed a hardware-centric strategy throughout its corporate history. The serial downturn of mainframe demand year-on-year forced IBM to confront a new business paradigm that past experience could not account for. IBM is facing 20 consecutive quarters of declining revenue[1]. Retail companies like Walmart and Macy’s have never previously encountered the threat of online commerce. Faced with a slow-growth market for switch & routers, Cisco is forced to seek alternatives. Under these circumstances, it’s no longer a matter of improving an enterprise within the sector, but to transition into different business models altogether.

 

 

Forecasting business outlook against historical trends can overlook material risks and opportunities. There is another branch of AI that enables business teams to manage the spectrum of information found in quarterly earnings call BI. This approach to AI augments human intelligence and actively works with humans to solve problems. It is based on man-machine interactions – bionic fusion – via visual feedback and expert inferences derived from instinct and automated meta-object discovery. For this reason, we believe bionic fusion can help executive teams create and grow a winning competitive edge.

 

Vision and Meta-Vision

Vision is one of the most efficient ways for living things to evaluate their surroundings for opportunity and risk. Computer vision mimics human vision to detect objects through image registration. Vision, however, does not have to be based on the physical view of an environment. Bats use sound and sonar to “see”. Scientists use radio signals and light spectrum, among other signals, to reconstruct the views of remote galaxies that may be several million light-years away. Similarly, Context Discriminant[2], an alternative approach to machine learning outlined in my previous post, applied on a body of textual information, maps a topological view of meta-objects that elevates domain expert insight discovery and execution. The resulting “Meta-Vision” helps business teams gain comprehensive insight across a wide spectrum of information from navigating market forces and assessing the competition to monitoring public relations media engineering. Applied to current events, Meta-Vision enables one to observe and triangulate the shift in public opinion over time. The ability to navigate risk and identify opportunity in qualitative information sources such as earnings call transcripts, paid media, earned media, voice of customer and investor sentiment enables business leaders to realize tactical and strategic execution.

 

The Meta-Vision Approach in predicting reaction to earning calls

Our AI development represents a necessary divergence from conventional statistical-based machine learning and rendering engines. As noted, the dynamic nature of business intelligence surrounding quarterly earnings calls severely limits the effectiveness of machine learning. Data may be new, volume may be low, terminology may not exist in the data dictionary or ontology. Context Discriminant technology enables business teams to discover and penetrate a wide spectrum of business intelligence. Meta-Vision enables a business team to create a topology where perceived strengths and weaknesses are manifested as “meta-objects”. Meta-object topology provides both visual cues into sentiment polarity and supporting facts for context and validation. Concerns can be addressed; head-winds, whether from direct competitors or market conditions, can be mitigated. Meta-Vision facilitates bionic fusion that help domain experts to navigate risk and opportunity in real-time.

 

Meta-Vision inputs can vary from current and relevant BI to earned media or “quarterly conference call” transcripts presented by most publicly traded companies. Predictive analytics on unstructured data (textual media) enable executive teams to better prepare their quarterly earnings call presentation in the following areas:

 

  • Identifying competitor strength & weakness
  • Qualifying company’s current strength & weakness
  • Historical analytics on prior earnings presentations and qualitative comparisons
  • Simulation of general public perception
  • Monitoring sector and/or industry meta-view trends
  • Monitoring the impact of macroeconomic and geopolitical events
  • Up-to-date qualitative supply and demand

 

Consequently, executive teams are able to proactively take actions to mitigate weaknesses and embolden areas of strength in preparation ahead of each conference call:

  • Embed presentation with tactical strategies that mitigate competitive headwinds
  • Maximize messaging optics (word choice, phrasing) to align with goals
  • Gain insight and clarity into the big picture – including external factors
  • Anticipate Q&A ahead of the event with objective simulation
  • Eliminate uneven distribution of weight due to subjectivity
  • Tactical strategic execution of vision
  • Continuous media monitoring for communication alignment for post-call outreach
  • Measure message pull-through, identify signals for misalignment, remedy and mitigate

 

Technology Validation

Context Discriminant and Meta-Vision have been deployed on earnings conference calls of DOW 30 and FTSE-100 companies with substantial results. Analysis was performed on quarterly earnings calls, earned media, and validated against analyst Question-and-Answers. The predictive insights drawn are consistent with market reaction to respective earnings calls. 


[1] IBM Revenue Declines for 20th Consecutive Quarter. Bloomberg. April 18, 2017. https://www.bloomberg.com/news/videos/2017-04-18/ibm-revenue-declines-for-20th-consecutive-quarter-video

[2] Introducing Context Discriminant – Artificial Intelligence for Tactical Strategic Execution with Bionic Fusion. http://www.datasciencecentral.com/profiles/blogs/introducing-context-discriminant-artificial-intelligence-for


Artificial Intelligence in Enterprise – Meta-Vision improves outlook and quarterly earnings call for publicly traded companies

Introducing Context Discriminant – Artificial Intelligence for Tactical Strategic Execution with Bionic Fusion

Introducing Context Discriminant – Artificial Intelligence for Tactical Strategic Execution with Bionic Fusion

Background
 

Prevailing AI technology for analytics prefer the use of statistical science as the foundation for machine learning (ML) on historical data to distill knowledge and experience. Whether it be supervised or unsupervised, the result is then incorporated into playback engines to analyze new data. These methods and procedures work well for predictable scenarios with known outcomes and known variables.

 

What if the variables are unknown, and the outcome is not predictable by past use-cases? When presented with this scenario, AI built upon the above premises will fail fast when compare to human experts. The “instinct” of a human expert – an art, rather than a science – enables them to adapt and discover the unknown. For this reason, it comes as no surprise that some human chess players can find ways to beat their machine counter-part.

 

The reality of textual analytics

 

With textual analytics, the objective will drive the solution preference. Conventional means of using ML on historical data takes on a different challenge. For example, if the objective is to understand sentiment, the goal is ranking the sentiment gradient. On the other hand, if the objective is user dialogue engagement, the goal is to analyze the question and to formulate an answer based on the question from a knowledge system. These applications are typically domain specific and will require a dictionary and ontology of the specific domain. However, if the purpose of textual analytics is to help a user gain insight into a dynamic range of subjects such as financial news across different industries, quarterly earnings reports and financial statements, customer reviews found in e-commerce sites, or media coverage of products and services then one may find most prevailing textual analytics inadequate. This is because the dynamic nature of subjects invalidates the above solutions. For this reason, my team turned to symbolic logic and propositional calculus to look for solutions.

 

Our quest for a new solution

 

The Internet has created a hyper converged digital world where business intelligence is everywhere. Media revelations that can impact the outcome of business decisions are streaming over high-speed pipes to business decision makers 24-7. Enterprise can no longer rely on traditional AI and ML to render solutions, equally, enterprise find it extremely difficult to find adequate human experts to simultaneously process voluminous business intelligence and devise Tactical Strategic Plans that can beat the competition and capture market opportunities. My team devised a solution to fuse the competitive advantages of both – to extend ML with symbolic logic and propositional logic so as to elevate the intelligence of human experts in solving complex problems. In the process, enterprise will be able to take advantage of the latest BI in advancing its business.

 

A new approach to artificial intelligence for textual analytics

 

While the details of this technology is beyond of scope of this writing, the general concept is not difficult to understand. We call this technology “Context Discriminant”. It is based on first order logic with symbols to infer, associate, prove or disprove premises using theorem-proving algorithms such as “resolution principle”. The idea behind this technology is to equip a software system with the ability to master a language such as English to the equivalent of a graduate student or researcher who can learn a core subject from a lecture or research medium. In this scenario, the medium uses English to introduce new subjects. In the process of knowledge transfer, the medium draws relationships between subjects and expresses the properties of the underlying context. The researcher, using English as a medium, can  learn any subject and acquire new knowledge by listening to lectures. In a similar manner, we implemented a software system to use the English language as the medium to learn any domain specific subjects in a set of documents. The software system uses visual charts to depict the discovered subjects, relationships, underlying context, properties, and references to source documents. When a user navigates through these properties, together with human thinking, it forms a bond of bionic fusion which enables the user to gain insights by drawing inference from these visuals.

 

The magic of Context Discriminant

 

My team developed this novel approach while searching for an automatic solution analyzing financial news. We have examined supervised and unsupervised machine learning in conjunction with financial news analytics and concluded that the pre-process and prerequisites of ML make it extremely difficult, if not impossible, to scale across various industries despite the commonality of business goals (i.e. supply & demand, competition, shareholder value, economy, outlook and revenue). Our technology evolved from a fully-automated solution into a universal tool that can be used to elevate the performance of any expert in any field. We draw parallels with contemporary man-machine fusion in bionic principles – information gathering, processing and optics such as those found in air traffic control radar or computerized tomography (CT) scans in medical applications. These devices gather information in real-time and provide continuous display to experts who can use it to perform a better job that otherwise would be impossible. We see this integration of mind and machine as the first step towards developing bionic systems elevating the capability and capacity of the human mind. Our system has been validated in the prediction of financial market forces and public company conference call transcripts. The results have been astounding. 


Introducing Context Discriminant – Artificial Intelligence for Tactical Strategic Execution with Bionic Fusion