IT for Enterprise AI

Next-Generation Data Strategies and Platforms for AI at Enterprise Scale

 

Significant opportunity for innovation and competitive advantage lies in applying AI to re-think how businesses operate and deliver dramatic improvements in how companies engage customers, make better use of their workforce and improve business operations. Business process re-engineering leveraging ubiquitous computing, data from corporate systems, the proliferation of sensors and the internet, and learning algorithms is commonly referred to as Digital Transformation. Yet today, prevalent use of AI in business is limited.

The use cases for AI in banking, for instance, are numerous. For example, AI applied to data produced by transaction and order systems, product systems, client masters, and document repositories can proactively identify the need to address corporate cash churn or to prioritize anti-money laundering efforts. AI and optimization techniques can be used to anticipate fluctuations in customer demand or supply disruptions, to better inform securities lending efforts, or for early identification of loan application risk. These are just a few of the many high-value AI use cases in the banking sector. Similarly, AI has been proven to significantly improve a wide range of processes across multiple industries, including manufacturing, aerospace, oil and gas, defense, healthcare, and utilities.

Continue reading or download the white paper.

Representative Value Chain: Banking

Banking Value Chain

The information technology challenges to delivering these AI applications are daunting.

The baseline capability required is aggregation and processing of rapidly growing petabyte-scale datasets continuously harvested from thousands of disparate legacy IT systems, internet sources, and multi-million sensor networks. In the case of one Fortune 500 manufacturer, the magnitude of the data aggregation problem is 50 petabytes fragmented across 5,000 systems representing customer, dealer, claims, ordering, pricing, product design, engineering, planning, manufacturing, control systems, accounting, human resources, logistics, and supplier systems fragmented by mergers and acquisitions, product lines, geographies and customer engagement channels (i.e., online, stores, call center, field).

 

Next Generation Application Platform Technical Requirements

The complete technical requirements for an AI application platform include:

  1. Data Aggregation from enterprise and extraprise systems, and sensor networks
  2. Multi-Cloud Computing for cost effective elastic scale-out compute and storage on private and public clouds
  3. Edge Computing for low-latency local processing and AI predictions/inferences
  4. Platform Services for continuous data processing, temporal and spatial processing, security and data persistence
  5. Enterprise Object Models to provide a consistent and comprehensive object model representation across a business
  6. Enterprise Microservices to provide a catalog of AI based software services
  7. Enterprise Data Security to provide robust user access authentication and authorization
  8. System Simulation using AI and Dynamic Optimization Algorithms with full lifecycle support including development, testing, and deployment
  9. Open Platform with support for multiple programming languages, standards-based interfaces (APIs), open source machine learning and deep learning libraries, and third-party data visualization tools
  10. Common Platform for Collaborative Development between Software Developers and Data Scientists

 

Cloud Vendor Tools and Structured Programming

A number of enterprises have attempted to address these requirements by assembling various native services and microservices offered by public cloud providers – e.g., AWS, Azure, and others – to build Enterprise AI applications. Leading vendors like AWS are developing a growing set of services and microservices, each of which provides useful functionality to address various Enterprise AI requirements. This requires the use of traditional “structured programming” to stitch together the various cloud services into a working application. While elastic cloud platforms such as AWS and Azure provide rich infrastructure-level services that are essential for Enterprise AI, this approach has proven ineffective for rapidly developing, deploying, and scaling Enterprise AI applications in a repeatable manner.

Figure: Structured programming results in complex code that is brittle and difficult and costly to maintain.

Complex Structured Programming

The problem with this approach is its high level of complexity: Because these systems lack a model-driven architecture like that described in the following section, developers need to employ structured programming to stitch together the various cloud services – resulting in numerous component interdependencies and the need to write, test, and debug many lines of code, creating brittle applications that are difficult and costly to maintain. Using traditional structured programming, the number of permutations of infrastructure service calls, enterprise systems and data integrations, enterprise data objects, sensor interfaces, application and data science programming languages and libraries to support application development is almost infinite.