GenAI – A Tale of Two Companies (Part 1)
I want to tell you a story. In a couple of years’ time it’ll be one of those ‘based-on-a-true-story’ stories. But it has yet to quite happen. But it will. You’re getting a sneak preview.
I’m going to take you to two islands; Cliste and Aireach. On each of these islands is the headquarters of a huge, multinational manufacturing company. One is called Farhatch, the other is called Ollimor.
Each controls a vast array of manufacturing facilities any one of which would leave you open-mouthed at its scale and complexity. These facilities are dotted across five continents; virtually the whole globe.
Many decades back, these were straight-forward manufacturing operations relying on pencil, paper and slide rule for design, a local workforce, short supply chains, a simple management structure and sales operations focused on domestic rather than international markets.
Over the years these two businesses have navigated political changes, economic crises, social changes, the transformation brought about by IT and the advent of the internet and later the cloud.
Through demand, innovation, acquisition and excellence, they grew to be the global entities they are today. Their size and scope mean that they’re already running complex, multi-cloud IT environments, and regrettably not every team is aware of what other teams are doing.
Now they stand on the cusp of a new era. A revolution. One heralded by the arrival of new cloud offerings and GenAI.
Both companies are fully aware of the power of AI to change their industry.
Their designers will be able to let their imaginations fly, supported and enhanced by AI’s ability to bring any idea, however fantastical, immediately to life. The AI will be able to pinpoint changes needed to differentiate products from competitors’ offerings.
The role of designers and engineers will evolve as AI puts the skills and knowledge of experts in other areas at their fingertips. So, as a product blueprint evolves, all the engineering parameters, component requirements and costs are automatically calculated.
Production managers will be able to streamline processes for maximum efficiency.
Their supply chain professionals can model demand for different products and their customised variations with extraordinary accuracy just from public pre-launch interactions with online visualisation tools. They can predict to the day and hour how many of which parts will be needed and where.
Every past and potential customer will be treated as a unique individual with sales teams able to deploy equally unique communications tailored to their tastes, needs and wants at will.
But beyond that AI will be able to take the users place and be a co-pilot, sending back an ocean of data that can be analysed and used in the design and sale of models yet undreamt of.
Improvements and updates will be continuous based on real time feedback.
AI will augment planning to the point where the company’s people will be able to prioritise colleagues, partners, and customers rather than spreadsheets. Perhaps counterintuitively, AI will allow huge businesses to rehumanise themselves.
The boardrooms at both island headquarters are increasingly focused on how they will use both cloud and AI. Both are aware that how they and their competitors use it may well determine which of them prevails in the battle for market share it will unleash.
But this is where the stories of the two companies’ stories diverge. In a future post I’ll take you to the Isle of Cliste to meet Farhatch, but for now let’s see how Ollimor is addressing the challenges posed by AI.
Ollimor is a great company. Its products are known across the world. One of the ways it has built global success is to devolve a lot of its decision making to its geographical divisions. There are six of them; Europe, North America, LatAm, Africa, MidEastIndia and Pacific.
Though the company has mostly rolled out the same tools across its IT estate, the divisions have a substantial say in what cloud they use and how. There is a logic in this. For regulatory reasons it’s necessary to have data stored within certain jurisdictions. Some cloud providers are better represented in some regions than others and thus offer better local failover capacity. Some providers are better suited to particular applications than others or offer different tools.
There have also been acquisitions that have come with legacy software and hosting arrangements. In addition, each division has invested in private cloud partly for compliance reasons and partly for security. It’s company policy to keep valuable IP off public cloud.
Responsibility for cloud spend at Ollimor falls to its Global Head of Cloud who in turn answers to the Head of Group Infrastructure and Services. But because of the company’s size and structure responsibility for gathering data and making decisions on cloud usage and costs is devolved to its regional heads of IT. The regional heads, in turn, rely on deputies in each individual territory.
Until now, and in a pre-AI world, it’s a set up that made perfect sense. It helped ensure decisions were taken as close to the ground as possible by people who understood the specifics of the business environments in which they worked. But, now the size and scale of the company’s cloud requirement, expanded by AI, have put this setup under huge pressure.
There have been attempts to centralise control of Ollimor’s IT but these met with some resistance and were not followed through.
There’s a fear that cloud could function as the thin end of a wedge for wider centralisation.
There is some internal politics. Local and regional teams are reluctant to give up what they see as their authority to make decisions on the ground. There’s a sense that the global IT operation, of which cloud is just a part, is simply too big to be controlled centrally and that doing so would just produce inertia just as the company needs to be at its most dynamic to take advantage of GenAI and the other new opportunities that cloud offers.
“Let’s focus on rolling out GenAI first and if we need to we can revisit control of the cloud later if we need to,” is how one executive put it.”
As things stand Ollimor is moving full steam ahead towards #GenAI adoption. However it’s already becoming apparent that country and regional IT teams are struggling to quantify their current and future cloud use. There’s insufficient visibility over local consumption and there is almost no visibility over global usage. That has serious ramifications for the deals that they’re striking with providers for public cloud.
The challenge grows as our story unfolds.
Meanwhile it isn’t proving easy to coordinate optimisation globally. There is a risk that the company’s different regions will take different path in their development and use of cloud resources. It’s also compromising efforts to streamline and to improve the efficiency of applications and transparency behind cloud costs.
However, for now, these concerns are taking a back seat to the excitement at the possibilities GenAI is opening up. Ollimor is dashing into the future with a determination to address any issues that emerge when their AI dream has been realised.
Meanwhile, as we’ll see over at Farhatch, things are proceeding quite differently.
To be continued… stay tuned, part two of our story will be posted next week.