Partner content: Celfocus’ strategy centres on Next-Gen Intelligence which combines AI, automation and human expertise, making work easier, faster and more valuable
In July 2025, the Massachusetts Institute of Technology (MIT) released its landmark report “The GenAI Divide: State of AI in Business 2025”.
Its message was clear: despite an estimated US $30–40 billion in global enterprise spending on Generative AI (GenAI), only around 5% of organisations are seeing measurable business returns.
The report exposes a widening gap – a GenAI Divide – between companies that experiment with AI and those that actually extract value from it. For most, GenAI remains a set of disconnected pilots. For others, it’s a foundation for measurable performance gains, cost optimisation and new business models.
Celfocus, as a system integrator specialising in data, AI and automation, stands at the heart of this transformation, helping clients move from proof-of-concepts to production, from experimentation to measurable impact. Celfocus follows a strategy centered on Next-Gen Intelligence – a concept that combines advanced AI, automation and human expertise to make work easier, faster and more valuable for both people and businesses.
Understanding the GenAI Divide
The MIT study examined over 300 AI initiatives and interviewed more than 150 senior executives across industries.
Its findings reveal a consistent pattern:
- High experimentation, low impact: Over 80% of companies have tested GenAI tools, yet only ~5% have scaled beyond pilots.
- Misplaced investment focus: Between 50% and 70% of GenAI budgets target front-office functions such as marketing or sales, while the highest returns appear in back-office automation, customer operations and business-process optimisation.
- A “learning gap”: The core barrier isn’t the technology itself but the lack of learning and adaptation in deployed systems. Too many solutions are static, poorly integrated into workflows and detached from data feedback loops.
- Value concentration: Sectors like Telecommunications and Media show tangible transformation, while industries such as Energy, Healthcare and Manufacturing are still in early, fragmented stages.
In short, the problem isn’t access to AI models – it’s how businesses apply and embed them.
Bridging Experimentation and Value: Celfocus’s Approach
Celfocus has long believed that AI only creates impact when it is deeply integrated into business workflows.
Through projects across Telecommunications, Financial Services, Utilities and Retail, Celfocus has demonstrated that sustainable GenAI transformation depends on a set of fundamental principles:
- Business-case validation before technology – Every GenAI initiative begins with measurable objectives such as reduced churn, improved efficiency or enhanced service experience, rather than technology enthusiasm.
- Workflow-level integration – Celfocus solutions are embedded in how people actually work. They connect to existing systems, guide human decision-making and automate routine actions.
- Continuous learning and adaptation – The most successful GenAI systems evolve. Celfocus designs feedback loops that enable AI models to learn from real outcomes, closing MIT’s identified “learning gap”.
- End-to-end governance – From data privacy to ethical use, Celfocus helps clients deploy GenAI responsibly, ensuring compliance and long-term trust.
- Scalable architectures – With reusable assets and modular design, Celfocus enables clients to move from pilot to enterprise scale without starting from scratch.
From Proof to Production: High-Value Use Cases
- Transforming Field Services to boost customer trust | Vodafone Group
Vodafone aimed to reduce repeat site visits and improve field service efficiency. Celfocus deployed Field Technician Assist, guiding engineers with real-time AI recommendations. The solution significantly improved first-time resolution, technician productivity, and customer trust.
A practical illustration of “crossing the divide” comes from Celfocus’s collaboration with Vodafone on the launch of Field Technician Assist, an AI-powered solution that is transforming Vodafone’s field service operations.
The solution combines Machine Learning with GenAI to analyse thousands of past interventions and technical documents, guiding field engineers with real-time, AI-driven recommendations on site.
The impact is measurable:
- One in four repeat visits eliminated,
- faster resolution times,
- and improved customer satisfaction.
These gains translate into more efficient operations, increased responsiveness without additional resources, and improved customer retention, directly contributing to churn reduction.
Know more about this project here.
- Reimagining Underwriting Through Intelligent Automation | Hiscox Europe
Hiscox Europe faced slow, inconsistent underwriting due to manual, document-heavy processes. Celfocus implemented an AI-powered workflow that automatically extracts and validates information, enabling 66% of requests to be processed fully automatically.
Celfocus partnered with Hiscox Europe to modernise underwriting, a core pillar of the insurance business, by transforming a heavily manual and document-intensive process into a unified, intelligent and scalable workflow powered by AI, GenAI and automation.
Underwriters handle a continuous flow of new business proposals from multiple brokers across different countries, arriving via emails, attachments and structured forms that reflect varied product rules and regulatory requirements.
Celfocus designed and delivered an end-to-end intelligent underwriting solution that removes friction from this process. Optical Character Recognition (OCR)-based document intelligence automatically interprets structured forms, while GenAI reads email bodies, identifies the insurance product and extracts the necessary fields.
The impact is measurable:
- 66% of underwriting requests processed fully automatically,
- 34% pre-populated, requiring only minimal validation from underwriters.
These results demonstrate a significant reduction in manual data handling, increased process consistency across geographies and greater operational agility. Underwriters can now focus on high-value activities such as risk evaluation and client advisory, while the cloud-native, microservices-based architecture ensures the solution can easily scale to new products, markets and underwriting volumes.
Know more about this project here.
- Streamlining Network Operations with Digital Twins & Agentic AI | BT
BT struggled to manage a highly complex network with fragmented tools and reactive processes. Celfocus deployed an integrated assurance platform using Digital Twins and Agentic AI to enable proactive network management. This has resulted in faster issue resolution, improved operational efficiency, and more reliable services.
BT faced challenges managing an increasingly complex network using multiple disconnected tools and manual processes. This made it harder to quickly understand what was happening on the network and how issues affected customers.
To overcome this, BT partnered with Celfocus and AWS to build an integrated platform that combines advanced analytics with agent-driven automation.
The platform uses multivariate KPI models and graph-based reasoning to detect anomalies early and perform rapid, cross-domain root-cause analysis. By linking network events to their impact on services and customers via explainable AI, the solution provides end-to-end anomaly detection, root-cause analysis and service impact insights – transforming BT’s 4G/5G CORE and RAN assurance capabilities from reactive to predictive and automated.
The impact is measurable:
- 100% reduction in L1/L2 Mean Time to Repair: AI-powered anomaly detection automates temporal anomaly scoring and accelerates issue identification.
- 50% improvement in proactive service management: The platform predicts performance degradation before SLA breaches, enabling proactive interventions.
- 30% increase in operational efficiency: Streamlined workflows and reduced reliance on domain-specific tools lower OPEX and minimise manual effort.
This clearer, end-to-end view of the network enables BT teams to act faster, reducing manual effort and preventing issues before they escalate. As a result, network operations become simpler, more proactive, and more cost-effective.
Know more about this project here.
These three Case Studies exemplify exactly what MIT identifies as the winning pattern: a narrow, high-value use case, tightly embedded in a workflow, continuously learning from feedback and delivering tangible business results.
The Path Forward: A Shift Towards Autonomy
As MIT notes, the next evolution of AI will be the Agentic Web – where autonomous, memory-based systems collaborate and act across workflows.
Unlike many organisations still experimenting with copilots, Celfocus has already operationalised Agentic AI, with systems that collaborate, optimise, and act based on intent rather than static instructions, transforming reactive operations into proactive, self-optimising ecosystems. These systems do not merely support humans – they work alongside them, augmenting decision-making, accelerating innovation, and enabling continuous optimisation at scale.
Celfocus’s Cognitive Automation solutions already deliver this, combining reasoning, contextual memory, and collaboration between multiple specialised agents.
The direction is clear: from assistance to autonomy, from tools to trusted agents, from data to decisions in motion.
Next-Gen Intelligence: How Celfocus is Building the Future
In light of MIT’s findings, it’s evident that the next leap for enterprises isn’t merely to adopt AI – it’s to make intelligence adaptive, human-centred and value-driven.
This is precisely the purpose behind Celfocus’s new strategic vision: Next-Gen Intelligence. This concept represents more than a technological shift; it’s a commitment to building intelligence from the inside out, combining advanced AI, automation and human expertise to empower faster, smarter decisions.
Where the MIT report exposes a GenAI Divide, Celfocus’s Next-Gen Intelligence strategy provides the bridge across it:
- Closing the learning gap – by designing systems that learn, adapt and act in real time.
- Amplifying human creativity – rather than replacing people, AI augments human talent, making work more purposeful.
- Data-driven scalability – treating data as a strategic asset, enabling intelligence to become accessible, replicable and reliable.
- Focus on real-world use cases – delivering solutions that tackle tangible challenges and generate measurable impact.
This shift turns enterprises into learning organisations, where intelligence continuously evolves and innovation becomes embedded in daily operations.
Conclusion: From Adoption to Advantage
The GenAI Divide reminds us that technology alone doesn’t create progress – application, learning and integration do.
Most organisations today are still experimenting with GenAI. A smaller group, however, is learning how to align it with business outcomes, workflows and culture. These are the organisations building a lasting competitive edge.
Celfocus’s Next-Gen Intelligence strategy marks the next stage of that journey, transforming AI adoption into enterprise evolution, and ensuring that intelligence is not just embedded in systems, but woven into the way people think, decide and create value.
Because the real power of AI lies not in the model, but in what it helps people and businesses achieve.
To know more, visit www.celfocus.com
About the author
Inês Rocha
Email: ines.rocha@celfocus.com
Job Title: Marketing Consultant at Celfocus
Inês Rocha is a Marketing Consultant at Celfocus with over seven years of experience in content strategy and creation. With a strong focus on Technology, Artificial Intelligence, and Digital Transformation, she specialises in crafting impactful narratives that inform, inspire, and support business impact. She holds a degree in Communication Sciences from NOVA University Lisbon and a postgraduate qualification in Storytelling.
