The second day of the simultaneous AI & Big Data Expo and Digital Transformation Week in London showed that the market is clearly in transition.
The initial excitement about generative models has faded. Enterprise leaders currently face friction in incorporating these tools into their current stack. The second day’s sessions focused less on large-scale language models and more on the infrastructure needed to run them (data lineage, observability, compliance).
Data maturity determines deployment success
The reliability of AI depends on the quality of the data. Northern Trust’s DP Indetkar warned against allowing AI to become “B-movie robots”. This scenario occurs when the algorithm fails due to insufficient input. Indetkar pointed out that the adoption of AI requires maturity of analytics. If your data strategy is not validated, automated decision-making will amplify errors rather than reduce them.
Eric Bobek of Just Eat supported this view. We discussed how data and machine learning can guide decision-making at the global enterprise level. Investments in the AI layer are wasted if the data foundation remains fragmented.
Kingfisher’s Mohsen Ghasempour also highlighted the need to turn raw data into actionable intelligence in real time. To reap the benefits, retail and logistics companies need to reduce the latency between data collection and insight generation.
Scaling in a regulated environment
In the financial, medical, and legal fields, there is little tolerance for error. Wiley’s Pascal Hetzscholdt addressed these areas directly.
Hetzscholdt said responsible AI in science, finance and law relies on accuracy, attribution and integrity. Enterprise systems in these areas require audit trails. Reputational damage and regulatory fines make “black box” implementation impossible.
Konstantina Kapetanidi from Visa outlined the challenges of building scalable generative AI applications using tools in multiple languages. Models are becoming active agents that perform tasks rather than just generating text. Allowing your model to use tools such as database queries creates a security vector that requires serious testing.
Parinita Kothari from Lloyds Banking Group detailed the requirements for deploying, scaling, monitoring and maintaining AI systems. Kothari challenged the idea of ”deploy and forget.” AI models require continuous monitoring, just like traditional software infrastructure.
Changes in developer workflow
Of course, AI is fundamentally changing the way we write code. A panel discussion with speakers from Valae, Charles River Labs, and Knight Frank explored how AI co-pilots are reshaping software creation. These tools speed up code generation, but also require developers to focus more on review and architecture.
This change requires new skills. A panel discussion with representatives from Microsoft, Lloyds, and Mastercard discussed the tools and mindset needed by future AI developers. A gap exists between the capabilities of today’s workforce and the needs of an AI-enhanced environment. Executives should plan training programs to ensure developers are able to fully validate the code generated by AI.
Dr. Gurpinder Dhillon from Senzing and Dr. Alexis Ego from Retool presented low-code and no-code strategies. Ego talked about using AI on low-code platforms to create production-ready internal apps. This method is intended to reduce the backlog of requests for internal tools.
Dillon argued that these strategies can speed up development without compromising quality. This suggests that for executives, internal software delivery becomes cheaper if governance protocols are maintained.
Employee abilities and specific utilities
A broader workforce is starting to collaborate with “digital colleagues.” EverWorker’s Austin Braham explained how agents are reimagining the workforce model. This term refers to the transition from passive software to active participant. Business leaders need to reevaluate human-machine interaction protocols.
Anthony Nolan’s Paul Airey gave an example of AI delivering literally life-changing value. He detailed how automation can improve donor matching and transplant scheduling for stem cell transplants. The utility of these technologies extends to lifesaving logistics.
A recurring theme throughout the event is that effective applications often solve very specific and high-friction problems, rather than aiming for one-size-fits-all solutions.
Manage migration
Sessions on the second day of the concurrent event demonstrate a shift in corporate focus to integration. The initial novelty has faded and been replaced by demands for uptime, security, and compliance. Innovation leaders need to assess which projects have the data infrastructure to survive contact with the real world.
Organizations should prioritize fundamental aspects of AI, such as cleaning data warehouses, establishing legal guardrails, and training staff to oversee automated agents. These details are the difference between a successful implementation and a failed pilot.
Executives should direct resources toward data engineering and governance frameworks. Without these, advanced models cannot provide value.
See also: AI Expo 2026 Day 1: Governance and data readiness enable the agent-based enterprise
Want to learn more about AI and big data from industry leaders? Check out the AI & Big Data Expos in Amsterdam, California, and London. This comprehensive event is part of TechEx and co-located with other major technology events such as Cyber Security & Cloud Expo. Click here for more information.
AI News is brought to you by TechForge Media. Learn about other upcoming enterprise technology events and webinars.
