Getting to grips with AI

On the 16th of May, 2018, Konvergent co-hosted an event with the Kier Group at their office in Fitzrovia, Central London. Our guest speaker, Chandan Rajah, Chief Big Data Architect and Director of Analytics and AI at various organisations, took invitees — including the Chief Architect at TUI Group, the Director of Architecture at Arup, the Group Head of Enterprise Architecture at Zurich Insurance and the Chief Architect at Nando’s — on a tour of current use cases for AI.


The case studies Chandan shared with us centre primarily around B2C online experiences—content and product recommendations, image analysis for metadata production and social media integration. However, what was apparent during his talk and the discussions following it is that AI is not only for the presentation layer or improving B2E interactions—for example, the Risk and Finance areas in some of our guests’ organisations are driving dozens of business cases to incorporate anomaly detection and predictive analytics in their operations.


Following the presentation, we gathered around our two round tables to enjoy a candid discussion about initiatives, opportunities, headaches and wins. We know our guests find these sessions therapeutic—it’s comforting to know someone else shares your frustrations, and there is always a lot of support and insightful advice doled out during the discussions.

Potential Use Cases

During our round table discussions, we uncovered the following list of use cases for AI—we thought it might be of interest for you when thinking about AI and opportunities to use it at your organisation:


  • Compliance
  • Fraud detection
  • Knowledge management
  • Natural language processing in internal and external communications
  • Service delivery improvements
  • Anomaly detection in internal processes (e.g. invoicing errors)
  • Contact risk/exposure analysis
  • Product strategy analysis
  • Product/search personalisation

The Konvergent View

Unlike many ‘flavour of the moment’ technologies, AI’s impact on business operations is a question of when, not if. Some organisations have already started investing in AI, but most are only at the POC/innovation stage right now with a few business cases making it through to implementation.


We think this delay in business demand is probably a good thing, as it allows the Chief Architect and the wider IT organisation to consider the business infrastructure that will be required to support AI. You could very easily unleash a monster on the business if the following considerations don’t form part of the overall adoption strategy:

Change Management.

Introducing AI into your business operations will cause major changes to your business processes. There must be appropriate impact assessments delivered to support AI tooling programmes to ensure your business capabilities are protected from interruption due to organisational chaos.


Ensuring the integrity of your underlying data structures is a must, but using model-based AI also exposes your organisation to risk of bias or errors in its customer-facing operations. To manage this wide range of risks and requirements, your business needs to assign overall responsibility and accountability for AI to the role of Chief Data Officer or another executive.

Human resources.

With changing business processes comes human resource impacts. This will likely mean some tasks will be accomplished much faster by a computer than by a human, so human effort will need to shift to more complex tasks and this may require new skillsets and training.

Technology needs.

Machine learning needs data. The data must be readily available and relatively current, as well as relevant. If you missed the big data bandwagon a few years ago, you’re going to need to plan your data capability now. For once, Konvergent are advising a technology-led roadmap rather than following the lead of the business strategy—when the business starts making demands it will be too late to start planning data lakes, so they need to be on your radar now.

Business needs.

Many technology organisations have invested in data science teams who are producing prototypes all over the place. But these AI endeavours will continue to be academic until there is focused business sponsorship driving them. Chief architects can help inspire the business to engage with AI and the transformational aspects of data analysis. Without this engagement, IT runs the risk of a lot of wasted innovation in POCs that deliver no business value.

More insights