Podcast: Tech Transformed podcast
Guest: John Newton, Chief Innovation Strategist at Hyland
Host: Dana Gardner, President and Principal Analyst at Interabor Solutions
Enterprise leaders rushing to integrate artificial intelligence (AI) into their operations often think the biggest challenge is the technology itself. In reality, the issue is much closer to home. It’s in the piles of unstructured enterprise data spread across documents, systems, and repositories.
In the recent episode of the Tech Transformed podcast, John Newton, Chief Innovation Strategist at Hyland, sits down with host Dana Gardner, President and Principal Analyst at Interabor Solutions. They discussed how enterprises can unlock the full value of enterprise AI by addressing fragmented information and building stronger governance frameworks.
Their conversation highlights that unstructured data is not an obstacle; it is the foundation for next-generation AI-driven productivity. As Newton stated, “The opportunity to truly use AI and use it effectively in your organisation really depends on that unstructured information.”
For companies looking to adopt AI on a large scale, the real work is in organising and contextualising their internal knowledge.
Is Unstructured Data the Hidden Fuel for Enterprise AI?
Most enterprise data does not sit neatly in structured databases. Instead, it exists in contracts, reports, emails, videos, policies, and operational documents, creating a vast amount of unstructured content.
The enormous amount of such unstructured data ends up creating a challenge for AI projects that rely solely on foundation models. Large language models (LLMs) may be trained on public data, but they cannot inherently access proprietary business intelligence.
Newton argued that enterprise AI must therefore be built around internal knowledge systems. “Foundation models can’t train on your internal information,” he explained. “What you really want is that information to be part of the AI when you’re answering questions, doing research, or executing business processes.”
This change requires organisations to rethink how information flows across the enterprise. Instead of isolated systems—CRM platforms, ERP databases, content repositories—companies need an interconnected information structure that connects multiple sources in real time.
Such a structure enables AI systems and AI agents to find the right data at the right time. This also improves decision-making, automation, and operational intelligence.
From Audits to Continuous Trust
Shift compliance from annual snapshots to embedded, automated controls that keep hybrid estates aligned with fast-moving global regulations.
How to Reorganise Chaotic Unstructured Data?
If unstructured data is the fuel, curation is the engine that drives effective AI. Newton emphasised that an enterprise data strategy must start with mapping, organising, and cleaning information assets. The aim is to reduce noise and increase clarity.
“I like to look at things from a signal-to-noise perspective,” Newton says. “Curation is the key to removing uncertainty in the information.”
The method could typically comprise a combination of several enterprise technologies, such as content management platforms with business process management (BPM) and AI agents and LLMs.
A pairing of the above strategies is aimed at helping enterprise data become more valuable. Enterprises can implement AI models to automate workflows, enhance knowledge discovery, and speed up processes across departments—from finance and manufacturing to customer operations.
Importantly, Newton noted that this work also allows flexibility in the AI ecosystem. With a solid information foundation, companies can use open-source models, hyperscaler services, or internal AI deployments without tying themselves to a single vendor.
In other words, an enterprise AI strategy should first focus on data readiness, not model selection.
Inside AI-First Supply Chains
Satellite, sensor and risk models converge into a frugal AI stack that translates climate signals into operational decisions at scale.
Key Takeaways
- Unstructured data is the foundation for effective enterprise AI.
- Data curation improves AI accuracy and reduces information noise.
- Connecting enterprise systems enables AI to deliver real-time insights.
- AI guardrails help manage security, compliance, and data governance.
- AI automation boosts employee productivity by reducing repetitive work.
Chapters
- 00:00 Unlocking AI's Potential with Unstructured Data
- 05:20 Signal to Noise: The Clarity Challenge
- 11:21 Guardrails for AI: Balancing Control and Flexibility
- 14:41 Harnessing the Enterprise Context Engine
- 17:48 Real-World Applications: Case Studies in AI
- 20:37 Curation: The Key to Effective Automation
- 22:21 Future Business Value: Productivity and Beyond
For more information, please visit hyland.com
To stay updated on B2B Tech front and centre, follow EM360Tech:
YouTube: @enterprisemanagement360
LinkedIn: @EM360Tech
X: @EM360Tech
Follow Hyland on all its major platforms:
YouTube: @HylandAI
LinkedIn: Hyland
X: @Hyland
#UnstructuredData #EnterpriseAI #DataCuration #AIGuardrails #LLMs #AIAutomation #FragmentedData #InformationManagement #SignalToNoise #EnterpriseContext #TechTransformedPodcast #Hyland #B2BTech
Comments ( 0 )