Enterprise Architectures for Artificial Intelligence (III)

Taxonomies of Artificial Intelligence

There are at least five ways to position AI in the enterprise landscape:

1. By processing method: batch, micro batch and real time
2. By algorithm type: pattern recognition, clustering, associations, scoring, predictive, classification, text, speech and image mining, …
3. By data type: high vs low dimensionality, graph data, self-describing data vs structured schema data, machine vs human sourced data, mediated data registration vs direct data registration,…
4. By data behaviour: volatile vs stable data values, long vs data persistency,
5. By analytics goals and or business process: churn prevention, prospect qualification, complex evaluations of loan applications, CVs, customer feedback, basket analysis, next best action proposals, fraud detection …

The enterprise architect will choose the relevant combinations between these taxonomies to produce a coherent end-to-end vision on the architecture. A possible selection criterion is the governance model used in the organisation. In a business monopoly analytics goals will be leading and combined with algorithm type. In an  IT monopoly processing methods combined with data behaviour is the most probable direction and in a duopoly, well,… that depends.

Let’s do an exercise and suppose this is the outcome of a duopoly governance model: combining the processing method with the algorithm type to indicate which processing method is most suited for the chosen algorithm type. Using this schema may help to manage expectations between the business and the IT people better.

	Batch	Micro batch	Real time
pattern recognition	Ideal method for large data sets	Suited for simple patterns	Only as a binary in/out of pattern decision which implies a large (batch) training set
clustering	Ideal method for large data sets	Suited for simple clustering criteria	Only as Y or N adherence to an existing cluster which implies a large (batch) training set
associations	Ideal method	Hardly possible	impossible
scoring	Develop a base line	Adjust the base line	Score against the base line
predictive	Develop a base line	Adjust the base line	Match with the trend
classification	Train the dataset	Classify new data	Simple classification
text mining	Train the dataset	Reveal polarity, topics, etc…	Deliver alerts
speech mining	Train the dataset	Reveal polarity, topics, etc…	Deliver alerts
Image & video mining	Train the dataset	Classify images	Deliver alerts

From this crosstab, it becomes possible to position the concrete algorithms, the data sets and their life cycle management, the ingestion volumes, timing and the technology to deliver on the various promises made.

Other methods will give you paths to the same end result: a coherent and methodical inventory of the landscape, linking business processes to AI and data mining initiatives and routines as well as the data and the applications to deliver the goods. Based on a gap analysis, the enterprise architect can develop a roadmap that communicates with all parties concerned.

Enterprise Architectures for Artificial Intelligence (I)

In the past three decades, I have seen artificial intelligence (AI) coming and going a couple of times. From studying MYCIN via speech technology in Flanders Language Valley to today’s machine learning and heuristics as used by Textgain from Antwerp University, the technology is here to stay this time.

Why? Because the cost of using AI has fallen dramatically not just in terms of hard and software but also in terms of acquiring the necessary knowledge to master the discipline.

Yet, most of the AI initiatives are still very much in the R&D phase or are used in limited scope. But here and there, e.g. in big (online) retail and telecommunications, AI is gaining traction on enterprise level.  And through APIs, open data and other initiatives, AI will become available for smaller organisations in the near future.

To make sure this effort has a maximum chance of success, CIOs need to embed this technology in an enterprise architecture covering all aspects: motivations, objectives, requirements and constraints, business processes, applications and data.

Being fully aware that I am trodding on uncharted territory, this article is –for now- my state of the art.

Introducing AI in the capability map

AI will enhance our capabilities in all areas of Treacy & Wiersema’s model, probably in a certain order. First comes operational excellence as processes and procedures are easier to describe, measure and monitor. Customer intimacy is the next frontier as the existing discipline of customer analytics lays the foundation for smarter interactions with customers and prospects.

The toughest challenge is in the realm of product leadership. This is an area where creativity is key to success. There is an approximation of creativity using what I call “property exploration” where a dimensional model of all possible properties of a product, a service, a marketing or production plan are mapped and an automatic cartesian product of all levels or degrees of each property with all the other properties is evaluated for cost and effectiveness. Sales pitch: if you want more information about this approach, contact us.

Capabilities where state of the art AI can play a significant role

Examples of capabilities where AI can play a defining role. Some of these capabilities are already well supported, to name a few: inventory management (automatic replenishment and dynamic storage), cycle time management (optimising man-machine interactions), quality management (visual inspection systems), churn management (churn prediction and avoidance in CRM systems), yield management (price, customer loyalty, revenue and capacity optimisation) and talent management (mining competences from CVs).

Areas where AI is coming of age: loyalty management and competitive intelligence, R & D management and product development.

In the next post I will discuss a generic architecture for AI in support of primary processes; Stay tuned and… share your insights on this topic!