Is 2025 the year we really move from GenAI hype to GenAI results? Recent research suggests yes, especially for the UK, which could see almost a doubling in economic growth over the next 15 years thanks to this cutting-edge technology.
But every technology leader knows they can’t predict every advance on the horizon, even while acknowledging their responsibility to plan for the future as much as possible. Across industries, leaders are faced with taking the plunge and investing in technologies like AI tools to future-proof their businesses. But without the right strategy and plan for adoption, you can end up drifting without a clear idea of where you’re going next.
Walking this tightrope takes a pragmatic approach that utilizes the best tools available while maintaining flexibility and control. Practical GenAI implementation is not about committing to one path. Instead, it’s about creating an AI ecosystem that adapts and evolves with your business needs. This may mean choosing platform-agnostic solutions to avoid vendor lock-in, embracing open source to benefit from flexibility and transparency, adopting hybrid and multi-cloud strategies to ensure the best environment for your AI workload, or focusing on the right size of your AI solutions.
CTO at Dell Technologies UK.
Pillars for practical GenAI implementation
Partnering with technology providers can ensure customers harness the power of AI—and manage the complexity, risk, and cost of diving into and supporting AI now and in the future. By offering flexible consumption models, an end-to-end AI-optimized IT infrastructure portfolio, an open ecosystem of deep partnerships with other leading AI companies, and a commitment to open standards, they can support a GenAI deployment that aligns with a company’s unique needs, risk tolerance and long-term vision. In short, they can help ensure a strategy that is not only ground-breaking, but also pragmatic and sustainable.
We can do that for our customers because of the lessons learned from our own AI journey. By implementing AI in our own operations, we have gained first-hand experience of its challenges and opportunities, giving us a deep understanding of what works and what doesn’t in the real world. Our “customer zero” approach, where we become our own first and best customer, ensures that our AI solutions are not just theoretical concepts, but grounded in practicality, refined through real-world experience and ready to deliver tangible results for our customers .
Through this practicality, we developed these five guiding principles to help you more quickly and efficiently implement AI technologies that will serve your business today and prepare you for your future business. These pillars of practical GenAI implementation are a testament to our own journey and our commitment to helping customers simplify complex technology.
1. Business data is your differentiator
Never lose sight of the fact that your data is a goldmine of insights, and unlike your competitors, you have exclusive access to it. You have a treasure trove of customer, operational and market data – information that reflects your company’s unique journey and expertise. This data is the secret to success in the AI race.
By building on pre-trained models and aligning them with your proprietary data, your differentiator, you can gain a competitive advantage through deeper customer insights (AI can analyze your customer data to uncover hidden patterns and predict future behavior), proactive risk management (AI) can detect fraudulent transactions in real-time by analyzing customer patterns and flagging anomalies) and improved decision-making (AI can analyze massive amounts of data to identify trends, predict demand and optimize pricing strategies – giving you the insights you need to make smarter and faster decisions).
2. Respect data gravity
Although data can be a treasure chest, it is never found in one pot. Data is highly distributed, with most of it residing on-premises, and more than 50% of enterprise data is generated at the edge.
For data to be effective, it must be close to applications and services that depend on it for efficient processing and analysis. It is better to give in to “data gravity” and bring AI to the data (the majority of which is on-prem) rather than moving enterprise data to available computing resources. Most organizations find it more efficient and effective to train and run AI models on-prem to minimize latency, lower costs and improve security. To turn data into actionable insights with AI, often in real-time, a combination of on-premise, edge and cloud deployments is essential. For this reason, 66% of UK decision makers prefer to build an on-prem or hybrid approach to the use and procurement of AI.
3. Right-size your AI infrastructure
There is no one-size-fits-all approach when it comes to AI. I’ve seen customers across multiple industries, in organizations of varying sizes, deploy their AI in countless ways – from locally on devices and at the edge all the way to massive hyperscale data centers. Not all models are large and not all AI workloads run in a data center. Or in the cloud. To avoid massive over- or under-provisioning, it will be important to right-size the AI solutions you deploy to your use case and requirements, so analyze your use cases and goals to determine the most appropriate infrastructure and model types.
4. Maintain an open, modular architecture
Equally important is to be aware that the AI landscape is constantly evolving and that no one can predict its future course. This means that a rigid, closed system can quickly become obsolete. Therefore, maintaining an open, modular architecture will be essential to help companies adapt to rapid changes in AI technologies and avoid being locked into outdated or inflexible architectures.
AI / GenAI workloads are a new class of workloads – requiring a new class of open, modern innovation that spans the entire AI estate: data layers and lakes, compute, networking, storage, data protection, and AI software applications. But it’s entirely plausible, if not likely, that new GPU infrastructure, algorithmic infrastructure, or inventions could emerge in the future that would require companies to adapt. The worst mistake you can make today is to invest in and commit to a closed, proprietary, one-dimensional AI system that is not flexible.
AI tools with open standards offer flexibility, transparency and a vibrant community for support and innovation. By integrating open standard solutions into their AI strategy, companies can avoid becoming dependent on a single vendor and adapt tools to meet their specific needs.
5. Create a thriving AI ecosystem
No single vendor can solve every AI challenge; collaboration is key. AI is a composite of many technologies, intellectual capabilities, and services that businesses will need to integrate with each other to succeed. Be sure to embrace vendors that enable an open ecosystem of partners, from big AI players like Microsoft to silicon providers like NVIDIA and Intel to open source leaders like Hugging Face.
Open ecosystems provide equal opportunities across the technology ecosystem, support the creation of new GenAI breakthroughs, and give customers greater access to innovation and flexibility. Access to open models and technologies can accelerate progress and solve problems worldwide, fueling a global “innovation engine” across all corners of industry, from individual developers and startups to the public sector and corporate organizations.
A real world approach to real world results
Navigating a new landscape successfully almost always requires a pragmatic approach that balances excitement with realism, preparation and careful execution. Being able to realize the value of new technologies requires the creation of strategic roadmaps, and when it comes to AI, the preparation, quality and storage needs of the data that feeds it have increased importance. Don’t get caught up in feeling like you have to turn yourself into an AI powerhouse overnight. Start by identifying a specific, achievable goal that has the capacity to generate business ROI, and power the path to success with a clear vision and the right partnerships.
We have compiled a list of the best free cloud storage.
This article was produced as part of TechRadarPro’s Expert Insights channel, where we feature the best and brightest minds in the tech industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing, you can read more here:
