Digital commerce teams rarely lack ideas. Most understand how AI, data, and personalisation could improve customer experiences. The problem, as explored in this episode of Don’t Panic, It’s Just Data, is turning those ideas into something that works at scale, in real time, and without slowing the business down.

Hosted by Dana Gardner, Principal Analyst at Interarbor Solutions, the discussion brings together Jürgen Obermann, Senior GTM Leader EMEA and Piotr Kobziakowski, Senior Principal Solutions Architect from Vespa.ai. Rather than focusing on hype, the conversation centres on the everyday realities of modern e-commerce systems and why progress often feels harder than it should.

AI introduces new expectations around speed, relevance, and adaptability. As a result, many digital commerce platforms are built on foundations designed for a different era. Years of development have resulted in fragmented environments, often based on microservices that once provided flexibility but now introduce complexity.

As Jürgen explains, even small changes can trigger long delivery cycles. Engineering teams may need months to safely update systems, not because the ideas are difficult, but because the infrastructure has become fragile.

Search is where most e-commerce journeys begin, yet many platforms still rely on keyword-focused approaches that struggle to interpret intent. Customers expect results that reflect who they are, what they want, and why they’re searching. Delivering meaningful personalisation requires systems that combine signals, context, and ranking logic in real time. Without that, experiences remain generic even when data is available.

The conversation then turns to architecture. Traditional search stacks, particularly Lucene-based systems, often hit performance limits when vector operations and advanced ranking are introduced. These capabilities tend to be bolted on rather than designed into the core. Piotr highlights a deeper issue, which is fragmentation. Search, ranking, recommendation, feature stores, and inference engines often live in separate systems. Each integration adds latency, duplicates data, and slows innovation.

This episode of Don’t Panic, It’s Just Data offers a calm, practical view of AI in digital commerce. Progress comes not from adding more complexity, but from simplifying how systems work together. When search, personalisation, and recommendation are designed as part of a cohesive whole, digital commerce platforms become easier to evolve and better equipped to serve both customers and the business.

For more insights into modern search architectures and AI-native commerce platforms, visit Vespa.ai.

• Many teams see the potential of AI, but face practical blockers.
• E-commerce companies struggle with operational, customer experience, and business challenges.
• AI technologies enable sophisticated personalised search experiences.
• Architectural bottlenecks often hinder e-commerce systems' performance.
• AI-native architectures can significantly improve search capabilities.
• Real-time personalisation is crucial for enhancing user experience.
• Separate systems for search and recommendations create inefficiencies.
• Phased migration is essential for transitioning from legacy systems.
• AI's impact on revenue can be profound when implemented effectively.
• Vespa is a comprehensive platform that integrates various functionalities.

00:00 Introduction to AI-Driven Search in E-Commerce

01:38 Challenges in Adopting AI for Digital Commerce

04:02 Architectural Bottlenecks in E-Commerce Systems

07:39 Designing AI-Native Search Architectures

12:00 Advancements in Personalisation for E-Commerce

16:21 Inefficiencies of Separate Search and Vector Systems

19:24 Phased Migration to AI-Native Platforms

21:51 Business Implications of AI in Search

23:57 Advice for Technical Leaders in E-Commerce

Vespa.ai is an AI search platform designed for building and operating large-scale, real-time applications. It brings together big data processing, vector search, machine-learned ranking, and real-time inference within a single system, enabling teams to deliver intelligent search, recommendation, and retrieval-augmented generation (RAG) at enterprise scale.

With native tensor support, Vespa allows complex ranking and decision logic to run directly in production, rather than being bolted on as separate services. This architecture reduces latency, simplifies system design, and makes it easier to evolve AI-driven applications as data, models, and business needs change.

In the recent episode of the Don’t Panic It’s Just Data podcast, Shubhangi Dua, Podcast Producer and B2B Tech journalist at EM360Tech, reports on the podcast shot live in London. Guest speakers, Pavel Dolezal, the CEO at Keboola, sit down with Vineta Bajaj, Group CFO, Holland & Barrett.

They get specific about how modern finance leaders move faster: start with one governed source of truth, then layer automation, and only then AI. They explore how the CFO role is evolving. From reporting numbers to also owning the non-financial “whys” behind them.

In the age of the AI boom, that shift turns every CFO into a product owner of data. But as Pavel Doležal puts it, without a clean, connected foundation, AI is just noise.

According to Vineta Bajaj, Group CFO of Holland & Barrett, the role of the CFO has fundamentally changed. Today’s CFO must act as a product owner for data, not just owning the numbers but also determining how data is defined, structured, and used throughout the business.

Drawing on her experience with Ocado Group, Rohlik Group (one of the fastest online grocery businesses in the world), and now Holland & Barrett, Bajaj points out that financial problems remain persistent across organisations.

Issues such as slow month-end closes, duplicated processes, delayed reporting, and limited decision-making speed are still common. These challenges are even greater in complex businesses that operate across multiple entities and countries. Differing charts of accounts, outsourced finance teams, and fragmented systems create added friction.

Bajaj stresses the answer isn’t "add another tool". CFOs should treat finance and data as a complete product, one that serves the business as its customer. This requires understanding finance processes, clearly defining financial and non-financial data, and prioritising what has the greatest impact on the business.

The Holland & Barrett CFO further emphasises that CFOs cannot pass this responsibility off to IT or BI teams. When data ownership is outside finance, it becomes someone else’s problem. However, when finance takes ownership of master data and its definitions while working closely with commercial and operational teams, it creates a single source of truth that the entire organisation can trust.

Also Watch: The Real Future of Data Isn’t AI — It’s Contextual Automation

Analytics, automation, and AI only work if the foundations are solid. Before adding AI assistants or real-time dashboards, CFOs must ensure that finance processes are clean, standardised, and automated. Poorly coded purchase orders, late journal entries, and inconsistent definitions can undermine even the most advanced technology.

At Holland & Barrett, this perspective led Bajaj to create a dedicated data function within finance. It ensures accountability for master data, definitions, and governance. The aim is not just to speed up reporting, but to gain deeper insights by linking financial outcomes with non-financial factors such as foot traffic, pricing, customer behaviour, and external influences like weather.

This integrated viewpoint allows finance teams to go beyond explaining variances and focus on the key business question: why performance changed and what happens next. It also opens up self-service analytics, reducing reliance on central BI teams and enabling decision-makers to act in real time.

Bajaj views AI as a powerful tool but not a shortcut. It prompts organisations to quickly address long-standing data and process issues. When data is well-defined and trusted, AI can facilitate scenario modelling, forecasting, and faster decision-making. Without proper discipline, AI merely adds to the confusion.

Ultimately, the future CFO must take an active role. They should engage with data, map out processes, ask difficult questions, and create a clear plan. Those who do will move faster than traditional finance models allow and help their organisations thrive in an AI-driven future.

1. The CFO’s role is evolving from reporter to product owner of data.
2. Slow month-end and fragmented processes block fast decision-making.
3. Finance must own data definitions to create a single source of truth.
4. Financial and non-financial data must be connected to explain the “why.”
5. AI only delivers value when financial data and processes are already clean.

1. 00:00 Introduction: The Modern CFO and Data
2. 01:05 What is the New Role of the CFO?
3. 03:34 The 3 Biggest Problems in Finance (Month-End, Reporting, Decisions)
4. 06:21 Why Every CFO is a Product Owner of Data
5. 09:37 Data Ownership: Should Master Data Sit in Finance or IT?
6. 13:30 The 3 Steps to Unleash Data Power (Process, Standardisation, Data Lake)
7. 17:08 How AI is Forcing Speed and Change in Finance
8. 20:25 The Future: Keboola's AI Assistant Roadmap
9. 21:36 Wrap-up and Final Thoughts

We live in a world where technology moves faster than most organisations can keep up. Every boardroom conversation, every team meeting, even casual watercooler chats now include discussions about AI. But here’s the truth: AI isn’t magic. Its promise is only as strong as the data that powers it. Without trust in your data, AI projects will be built on shaky ground.

In this episode of Don’t Panic, It’s Just Data podcast, Amy Horowitz, Group Vice President of Solution Specialist Sales and Business Development at Informatica, joins moderator Kevin Petrie, VP of Research at BARC, to tackle one of the most pressing topics in enterprise technology today: the role of trusted data in driving responsible AI. Their discussion goes beyond buzzwords to focus on actionable insights for organisations aiming to scale AI with confidence.

Amy opens the conversation with a simple but powerful observation: “No longer is it okay to just have okay data.” This sets the stage for understanding that AI’s potential is only as strong as the data that feeds it. Responsible AI isn’t just about implementing the latest algorithms; it’s about embedding ethical and governance principles into every stage of AI development, starting with data quality.

Kevin and Amy emphasise that organisations must look at data not as a byproduct, but as a foundational asset. Without reliable, well-governed data, even the most advanced AI initiatives risk delivering inaccurate, biased, or ineffective outcomes.

Responsible AI is more than compliance or policy checkboxes. As Amy explains, it is a framework of principles that guide the design, development, deployment, and use of AI. At its core, it is about building trust, ensuring AI systems empower organisations and stakeholders while minimising unintended consequences. Responsible data governance is the practical arm of responsible AI. It involves establishing policies, controls, and processes to ensure that data is accurate, complete, consistent, and auditable.

The takeaway from this episode is clear and that is responsible AI starts with responsible data. For organisations looking to harness AI effectively:

1. Invest in data quality and governance — it is the foundation of all AI initiatives.
2. Embed ethical and legal principles in every stage of AI development.
3. Enable collaboration across teams to ensure transparency, accountability, and usability.
4. Start small, prove value, and scale — responsible AI is built step by step.

Amy Horowitz’s insight resonates beyond the tech team: “Everyone’s ready for AI — except their data.” It’s a reminder that AI success begins not with the algorithms, but with the trustworthiness and governance of the data powering them.

For more insights, visit Informatica.

• AI is only as good as its data inputs.
• Data quality has become the number one obstacle to AI success.
• Organisations must start small and find use cases for data governance.
• Hallucinations in AI models highlight the need for vigilant data oversight.
• Reputational damage from AI failures can be severe for organisations.
• Metadata plays a crucial role in data management and governance.
• Collaboration between data, AI, and development teams is essential.
• Data governance is a must-have, not a nice-to-have.
• Organisations need to enable their lines of business for effective AI implementation.
• Everyone is ready for AI, except for the quality of their data.

00:00 The Importance of Responsible AI and Trusted Data

02:49 Defining Responsible AI and Data Governance

05:40 Challenges in Data Quality and Governance

08:51 Real-World Examples of Data Quality Issues

11:51 The Role of Employees in Data Governance

14:41 Successful AI Outcomes Through Responsible Data Practices

17:42 The Risks of AI Governance and Reputational Damage

20:42 Collaboration Across Data, AI, and Development Teams

23:34 The Future of Metadata and Data Management

26:42 Key Takeaways for Data and AI Leaders

Informatica, founded in 1993, is an enterprise data management company headquartered in Redwood City, California. The company provides software products for data integration, data quality, master data management, and data governance. With approximately 9,000 global customers across various industries, Informatica has positioned itself as a significant player in the data management market.

“What is the true value of our data and AI initiatives?” 

Too often, we drive all our energy into tools, processes, and outputs, but forget to ask ourselves how what we build actually makes a difference. For enterprises, this means looking beyond AI models and dashboards to see how our data drives real, measurable impact. Understanding the difference between output and outcome is what separates activity from transformation.

In this episode of Don’t Panic, it's Just Data, host Doug Laney and Nadiem von Heydebrand, CEO and Co-founder of Mindfuel, explore how organisations can turn data and AI efforts into actionable business outcomes. They discuss the concept of the “value layer”, a framework connecting data initiatives to business needs, emphasising the importance of understanding business problems before developing solutions.

Nadiem stresses that prioritising initiatives and fostering strong collaboration between business and data teams are critical to unlocking maximum value from data and AI efforts.

Many organisations are investing heavily in data and AI, but turning these investments into real business value remains a challenge. This is because a critical gap exists between technical execution and business outcomes. Data and AI teams work on initiatives without first clarifying what business problems they're solving or how success will be measured.

Data and AI Impact Management bridges this gap by establishing the “value layer" between business strategy and technical platforms. This approach starts with structured demand management for use cases, enables systematic prioritisation based on actual value potential, and tracks initiatives throughout their lifecycle to ensure they deliver impact against business goals. This shift, from building solutions in search of problems to solving qualified business problems with purpose-built solutions, transforms data and AI teams from technical support functions into strategic partners who deliver value, stronger strategic alignment, and lasting competitive advantage. 

Nadiem says, “Applying a product mindset within data initiatives is key, and it's the foundational effort to be able to drive value.” 

He also notes that not every use case delivers direct financial impact, and the value layer helps clarify demand, manage use cases effectively, and uncover each initiative’s business value

For more insights and solutions, visit Mindfuel

• Organisations struggle to connect data initiatives to business outcomes.
• The value layer is essential for linking data to business demands.
• Understanding the actual business problem is crucial for success.
• Value management encompasses the entire lifecycle of initiatives.
• A product mindset helps focus on outcomes rather than outputs.
• Not all data use cases have direct dollar values.
• Data and AI impact management creates transparency for data teams.
• Establishing a product mindset is key for data products.
• Connecting processes to the operating model enhances effectiveness.
• Collaboration between business and data teams is vital for unlocking value.

00:31 Introduction: Don't Panic, It's Just Data

01:37 The Missing Piece: Introducing the Value Layer 

07:11 Value Management Lifecycle

10:46 Product Mindset in Data Initiatives

14:10 Distinguishing Value and Impact 

17:04 Impact Management and Investment Justification 

19:34 Mindfuel's Three-Step Guide to Impact Management 

21:00 Conclusion and Key Takeaways

Mindfuel is a data and AI impact management platform that gives data, analytics and AI teams a single source of truth to prioritise high-impact use cases, connect initiatives to business outcomes, and demonstrate ROAI. It replaces scattered tools and reactive, manual processes with a structured approach to managing use cases and data and AI products. This enables organisations to reduce business case bias, eliminate inefficiencies, and clearly communicate the value of AI initiatives, driving enterprise-wide trust, transparency, and impact. 

As AI becomes a central pillar of business decision-making, enterprises face a new challenge, and that is making their data AI-ready. It’s no longer enough to collect and digitise information. For organisations, data must be structured, contextualised, discoverable, and usable—both by humans and intelligent systems.

AI can only deliver if your data is truly ready, but most enterprises are drowning in fragmented, incomplete, or slow-to-update data. In this episode of Don't Panic, It's Just Data, host Doug Laney and Sushant Rai, Vice President of Product of AI and Data Strategy at Relito, explore how modern data unification strategies are changing enterprises, enabling AI to deliver faster, more reliable insights. They focus on the shift from traditional Master Data Management (MDM) to next-generation AI-ready data cores, uncovering the risks of fragmented data and the strategies to overcome them.

AI, especially large language models (LLMs), is changing how people interact with data. Analysts, executives, and frontline teams now expect natural language queries and instant, actionable insights.

Sushant explains:

When you embed intelligence into your data layer, AI can help you manage and scale your data without drowning your teams in manual work. This will only work if your data is structured, clean, governed, and constantly updated, everything that makes it truly AI-ready.

The volume of data being turned over daily is staggering. 

As Sushant notes:

This gap is driving the change in the modern data platforms, where AI automates stewardship, enriches data continuously, detects anomalies, and maintains quality in real time.

Want to learn more about modern data unification and AI-ready platforms? Visit Reltio.com for insights, resources, and case studies.

• Data unification provides a trusted, real-time view of key business elements.
• Organizations must balance speed and trust in data management.
• Classic MDM is evolving into modern data unification platforms.
• Real-time data access is crucial for AI and analytics.
• AI can enhance data quality and governance processes.
• Successful data initiatives require clear business outcomes and ownership.
• Data unification should be viewed as a business platform, not just an IT project.
• AI agents will play a significant role in automating data governance.
• Organizations need to focus on both structured and unstructured data.
• The future of data management involves continuous unification and enrichment of data.

00:00 Introduction to Data Unification and AI

07:52 The Importance of Data Unification in Enterprises

15:44 AI and Data Quality Management

23:20 Organizational Success Factors for Data Initiatives

25:16 Future Trends in Data and AI

At Reltio, we believe data should fuel your success in the enterprise AI era. Reltio Data Cloud™ is the agentic data fabric for the enterprise—powering real-time data intelligence and AI transformation. Reltio’s cloud-native SaaS platform delivers unified, trusted, and context-rich data across domains in real time. With Reltio, organizations gain 360-degree views of customers, products, suppliers, and more—mobilized in milliseconds to any application, user, or AI agent. Trusted by the world’s largest enterprises across life sciences, financial services, healthcare, technology, and more, we help organizations fuel frictionless operations, drive innovation, and reduce risk.

While the role of a chief data officers (CDOs) was traditionally focused on regulatory compliance, it has now expanded to empowering the consistent and effective use of data across organizations to improve business outcomes. One of the most effective ways for CDOs to demonstrate their value is by developing a data strategy that is closely aligned with business goals, processes, and outcomes. 

In the latest episode of Tech Transformed, host Kevin Petrie, VP of Research at BARC, speaks with Brett Roscoe, Senior Vice President and GM of Cloud Data Governance and Cloud Ops at Informatica, about the evolving role of CDOs. Their conversation explores how CDOs are transitioning from data stewards to strategic leaders, the importance of data governance, and the challenges of managing unstructured data.

As Roscoe notes, “CDOs are now pivotal in AI strategy,” reflecting how the role has grown from compliance oversight to guiding enterprise initiatives that directly support organizational goals.

In this day and age, CDOs are tasked with ensuring that data is both accessible and reliable, providing a foundation for informed decision-making across business units. This includes establishing policies for data quality, access, and governance, which Roscoe highlights as essential: “data governance is foundational for AI.” At the same time, unstructured data ranging from documents and emails to multimedia adds complexity that requires careful management to make it useful while minimizing risk. “Unstructured data presents challenges,” he adds, emphasizing the need for structured oversight to fully leverage these assets.

Although technology and analytics are evolving rapidly, the CDO’s role in aligning data with strategic initiatives is critical. By connecting data assets to business processes, CDOs help ensure that initiatives are informed by reliable, well-governed information and can deliver measurable results.

For anyone looking to understand the evolving responsibilities of CDOs, the importance of governance, and strategies for handling unstructured data, this episode of Tech Transformed provides a detailed and practical discussion.

For more insights, follow Informatica:

• X: @informatica
• Instagram: @informaticacorp
• Facebook: https://www.facebook.com/InformaticaLLC/
• LinkedIn: https://www.linkedin.com/company/informatica/

• CDOs are now central to shaping AI strategies and driving business growth.
• Robust data governance is crucial for the successful deployment of AI technologies.
• Unstructured data presents unique challenges and opportunities for AI development.
• A balance between centralized governance and federated operations is essential.
• Securing executive support is vital for the success of CDO-led initiatives.
• Engaging business stakeholders enhances the impact of AI projects.
• Demonstrating ROI through clear metrics is key to sustaining AI investments.
• AI governance must extend beyond data to include models and agents.
• New measures are needed to ensure the quality and governance of unstructured data.
• CDOs must navigate the tension between fostering innovation and maintaining governance standards.

00:00:00 Introduction to the Podcast and Guests

00:03:00 Brett Roscoe's Background and Role

00:06:00 The Evolving Role of CDOs

00:09:00 Data Governance as a Foundation for AI

00:12:00 Challenges with Unstructured Data

00:15:00 Governance Frameworks for AI and Data

00:18:00 Centralization vs. Decentralization in Data Governance

00:21:00 CDO Strategies for Success

00:24:00 Conclusion and Future Outlook

Informatica, founded in 1993, is an enterprise data management company headquartered in Redwood City, California. The company provides software products for data integration, data quality, master data management, and data governance. With approximately 9,000 global customers across various industries, Informatica has positioned itself as a significant player in the data management market.

The challenge all organisations, big and small, face is answering and implementing solutions to solve this key question: How can finance and accounting teams work faster, smarter and more accurately?

In the recent episode of the Don’t Panic It’s Just Data podcast, host Scott Taylor, The Data Whisperer and Principal Consultant at MetaMeta Consulting, speaks with Kevin Gibson, CPA and Principal Solutions Engineer at insightsoftware. They talk about the constantly changing nature of financial reporting. 

Additionally, they discuss the pros and cons of modern financial reporting and the importance of connecting financial data with familiar tools like Excel. The conversation also touches on the future of financial reporting technology and the need for organisations to adapt to changing data access needs.

“With all this uncertainty, companies are being asked to look at their data in different ways. They want to pivot it, slice it, and dice it,” Gibson tells Taylor, encapsulating the theme of this episode. “They’re being told to do more with the data — what does it mean, how do we read it, how do we understand it, how do we analyse it?”

The issue is that, as enterprises invest in digital transformation, finance teams struggle most with limited access to the data they need to support their analysis.

“The ideal state,” Gibson adds, “would be: I can get what I want, when I want, and how I want it — without asking questions. But let’s be honest — that doesn’t exist today.”

However, the good news is the data exists, Gibson says. The ugly part is that organisations can’t get to it. Many of the data accessibility issues have been attributed to cloud migration. 

“When you move your data to the cloud, you think: it’s cheaper, it’s more secure, it’s easier to maintain. But here’s the problem: you don’t control it anymore. Some cloud providers make access difficult or costly. So finance teams feel stuck,” he explains.

Also Watch: Stop Fighting Excel: How to Turn Your Spreadsheets into a Real-Time Reporting Powerhouse?

For decades, finance professionals have relied on Excel, which Gibson refers to as the “largest data warehouse in the world.” “There are 1.1 billion users of Excel today,” he says. “And let’s be honest, I haven’t met an accountant yet who says they hate it.” 

Finance prospers in Excel, but IT often views it as a risk. This leads to a constant back-and-forth between usability and control. Gibson believes that the solution is to equip both sides- finance and IT with real-time, governed data inside Excel. 

That’s where insightsoftware comes in. “We can connect directly to these systems and give finance teams back their real-time access — not just to pieces of data, but all of it,” says Gibson. “Literally every piece of data can be accessed.” 

With tools like Spreadsheet Server, finance professionals can work in Excel — their “comfort food,” as Gibson calls it — while drawing directly from live ERP data in the cloud. 

“We give them insight — that’s what our software does. It gives them visibility into their data. Excel isn’t going away, and our job is to make it work even better.”

To learn more, watch or listen to the podcast on EM360Tech.

Also Watch: Struggling with ERP Data? How to Get Real-Time Reporting in Excel

• Finance professionals are facing increased pressure to analyse data amidst uncertainty.
• The ideal state for finance teams is immediate access to data without barriers.
• Modern financial reporting has its good, bad, and ugly aspects, primarily revolving around data accessibility.
• Different industries have unique data needs and KPIs that impact financial reporting.
• Excel remains a critical tool for finance professionals despite the rise of cloud-based solutions.
• Organisations must find ways to connect their financial data with familiar reporting tools.
• The future of financial reporting will continue to evolve around Excel and data accessibility.
• Companies are beginning to realise the limitations of cloud-based systems and may revert to on-prem solutions.
• AI is emerging as a significant factor in data access and reporting.
• C-level executives should engage with finance teams to ensure they have the tools they need.

• 00:00 Introduction to the Podcast and Guest
• 00:59 Understanding the Role of Finance in Uncertain Times
• 04:00 The Good, Bad, and Ugly of Financial Reporting
• 08:13 Types of Data and Industry-Specific Challenges
• 10:00 Connecting Financial Data with Reporting Tools
• 13:59 The Future of Financial Reporting and Technology
• 17:58 Key Takeaways for C-Level Executives

"A flaw of warehouses is that you need to move all your data into them so you can keep it going, and for a lot of organisations that's a big hassle,” says Will Martin, EMEA Evangelist at Dremio. “It can take a long time, it can be expensive, and you ultimately can end up ripping up processes that are there."

In this episode of the Don’t Panic It’s Just Data podcast, recorded live at Big Data LDN (BDL) 2025, Will Martin, EMEA Evangelist at Dremio, joins Shubhangi Dua, Podcast Host and Tech Journalist at EM360Tech. They talk about how enterprises can enable the Agentic AI Lakehouse on Apache Iceberg and why query performance is critical for efficient data analysis. 

"If you have a data silo, it exists for a reason—something's feeding information to it. You usually have other processes feeding off of it. So if you shift all that to a warehouse, it disrupts a lot of your business," Martin tells Dua. 

This is where a lakehouse comes into play. Organisations can federate their access through a lakehouse data approach. They can centralise access to the respective organisation’s lakehouse while keeping their data in its original location. Such a system helps people get started quickly.

In terms of data quality, if you access everything from one location, even with separate data silos, you can see all your data. This visibility allows you to identify issues, address them, and enhance your data quality. That’s beneficial for AI, too, Martin explains. 

Lakehouse has been recognised for unifying and simplifying governance. An imperative feature of a lakehouse is the data catalogue, which helps an organisation browse and find information. It also secures access and manages permissions.

"You can access in one place, but you can do all your security and permissions in one place rather than all these individual systems, which is great if you work in IT,” reflects Martin. "There are some drawbacks to lakehouses. So, a big component of a lakehouse is metadata. It can be quite big, and it needs managing. Certain companies and vendors are trying to deal with that."

With AI and AI agents, it’s become even harder to optimise analytics on a lakehouse. However, this has been improved as technical barriers are disappearing. Martin explains that anyone can prompt a question; for instance, an enterprise CEO could ask questions about the data and demand justifications directly. 

In the past, a request would have to be submitted, and then a data scientist or engineer would create the dataset and hand it over. Now, engineers' roles have changed to focus on better optimisation. They help queries run smoothly and ensure tables are efficient. Agents cannot assist with that.

Also Listen: Dremio: The State of the Data Lakehouse

Vendors such as Dremio provide services to manage and optimise lakehouses. They offer autonomous features to help set up the workflow. Martin says that in many cases, Dremio learns from the clients’ actions and improves their system. “This is evident in our reflections, which are optimised datasets that speed up performance,” he added. 

“In other situations, we handle tasks like file compaction and garbage collection, which are often less exciting for engineers. Now, there’s no need for engineers to manage those tasks, which benefits everyone.”

As a lakehouse provider, Dremio is Iceberg native. They began their journey as a lakehouse provider and continue down this road. Now the industry has shifted gears to focus on lakehouses too, first with Snowflake and now even Databricks, which has developed its format with Delta Lake.

The ultimate goal is to incorporate more features—permissions, governance, and fine-grained access control. “These capabilities are things vendors typically sell, but they will soon become widely available for free,” Martin tells Dua.

Learn More: Visit dremio.com for more information on open data lakehouse technology.

• Agentic AI and Apache Iceberg are current hot topics.
• Lakehouses offer quicker, less disruptive data access for AI compared to data warehouses.
• Centralised access in a lakehouse improves data quality and simplifies AI integration.
• Lakehouses, with their data catalogues, ease governance and permission management for AI agents working with sensitive data.
• Apache Iceberg is resolving metadata format issues, though metadata management remains an overhead.
• Dremio, an Iceberg-native provider, champions open source and interoperability, offering autonomous optimisation features to free engineers from mundane tasks.
• Beyond technology, a robust data strategy is crucial for organisational data improvement.
• Agentic AI will evolve to handle more delegated, multi-step tasks with less supervision.
• The open-source ecosystem will see consolidation and improved features, making advanced catalogue and governance tools widely available.
• Ultimately, for IT decision-makers, the quality of data is paramount for all analytical endeavours, including AI.

• 0:00 - Introduction to Agentic AI
• 0:35 - Discussing Big Data London, Hot Topics: Agentic AI and Apache Iceberg
• 1:37 - Data Lakehouse vs. Data Warehouse for AI
• 2:30 - Data Quality and AI with a Lakehouse
• 3:18 - AI Agents and Sensitive Data: Governance with a Lakehouse
• 4:19 - Challenges and Solutions in Lakehouse Technology (Apache Iceberg)
• 5:47 - Dremio's Use Cases and Interoperability
• 7:40 - Dremio's Standout Features and Autonomous Optimisation
• 9:39 - The Importance of Data Strategy
• 10:29 - Future of Agentic AI
• 11:34 - Future of the Open-Source Ecosystem
• 12:51 - Final Takeaway for IT Decision Makers: Data Quality is Critical
• 13:51 - Conclusion

At Big Data LDN (BDL) 2025, Keboola CEO Pavel Dolezal presented a new data agent designed for all business users, not just engineers. With a mission to make AI, automation, and data easy to access, relevant, and useful across the organisation, Dolezal revealed that the data agent has been embedded with contextual intelligence and generative AI.

“While we typically assist data engineers with building the pipeline, we took the same data agent and built a different environment for it — a chat-like environment. By default, the chat has context, knows what to do, knows where not to go,” the Keboola CEO unveiled on the Don’t Panic It’s Just Data podcast. 

In the EM360Tech podcast recorded live at BDL, Dolezal spoke to Christina Stathopoulos, the Founder of Dare to Data in the recent episode of the Don’t Panic It’s Just Data podcast. They talked about the new Keboola Data Agent and it plays a key role in AI-backed change and the growth of large language models (LLMs) in business.

“Anyone can be technical now. It’s context that matters,” stated Dolezal, also the co-founder of Keboola. He presented a strong argument for why enterprise data strategies are falling short and how a new wave of smart tools will change that.

“The pipeline of what you can do is limitless if you build it for people in business,” he added. “You can't keep data and AI just in the hands of engineers anymore. That model doesn’t scale.”

As businesses face a growing number of data sources — sometimes over 300 SaaS platforms and more than 80 departments — managing, governing, and activating that data has become a challenge. The appealing promise of AI often adds another layer of complexity.

Enterprise leaders were told that AI would simplify data workflows. Instead, many found themselves managing disconnected tools and failed pilot projects.

“We all read the MIT study. 95 per cent of AI proofs of concept don’t make it to production,” the CEO of Keboola highlights. “Why? Because large language models (LLMs) need context. And enterprise data environments are anything but simple.”

At Keboola, context is crucial, he emphasised. It includes not just metadata but also event logs, debug trails, and orchestration details, including the complete story behind every data product.

“LLMs thrive on context. The more relevant context you provide, the better the outcome. But in today’s data stack, where your context is spread across 15 tools, that's nearly impossible.”

This is where Keboola’s new Data Agent comes in. It is a generative AI interface built directly into data workflows, capable of understanding and acting on both the structure and state of a company’s data.

Watch the podcast for further insights on EM360Tech. 

• Focus on context and domain knowledge rather than technical skills.
• Ease of acquiring technical skills compared to the importance of understanding business processes.
• Agents will run business processes both internally and externally.
• Need for infrastructure to support agents 
• Challenges of provisioning ad hoc environments
• Use agents to automate business processes.
• Importance of data governance.

• Introduction & Keboola's Mission: 0:00
• Challenges in Modern Data Management: 2:23
• Impact of Complexity on Teams: 4:04
• LLMs in Data: Potential and Pitfalls: 5:50
• Keboola Data Agent: Bridging the Gap: 8:12
• Evolution of Keboola Data Agent: 13:00
• Future Vision for LLMs & Agents: 15:05
• Key Takeaway for Leaders: 17:12

With an ever-changing business climate, companies have begun to shift their focus to unstructured data. In the past, unstructured data was challenging to deal with, considering the volume, governance and compliance, so organisations mainly focused on structured datasets. 

However, with the rise of generative AI and large language models (LLMs), Reece Williams Griffiths, Field CTO of Collibra, says that we can no longer overlook 80 percent of enterprise content—from transcripts and PDFs to emails and images.

In this episode of the Don't Panic It's Just Data podcast, host John Santaferraro, CEO and Head Research Analyst at Ferraro Consulting, talks with Griffiths, also Co-Founder and CEO of Deasy Labs (acquired by Collibra). They also talk about the change brought to Collibra after acquiring Deasy Labs. 

Following Collibra’s acquisition of Griffiths firm, Deasy Labs, he explains how this merger is making AI truly achievable for businesses. Deasy became renowned for its goal of simplifying data preparation. With Collibra, it’s leading the development of the tools necessary to create order from the chaos and build a unified AI enterprise.

Together, they created the first unified governance and catalogue platform for both structured and unstructured data. This single-hub approach is vital for a future where AI agents treat all data types equally.

Griffiths tells Santaferraro that, historically, Collibra, like others, focused only on structured data. Now, by combining Deasy’s capabilities, the platform provides a single entry point and a smooth experience for all data assets.

One outcome of a unified data strategy is simplified AI use cases. Since AI applications often need to access both tabular data (structured) and documents (unstructured) to give complete answers, unification offers the necessary routing and flexibility, the Field CTO explains.

To effectively use a huge quantity of unstructured content, it must be prepared. Griffiths describes a four-layer data preparation funnel that goes beyond simple classification to deep semantic embedding, ultimately creating a Knowledge Product.

The talk of the moment is the knowledge data product, which the Collibra speaker says is familiar in the structured data scenario; however, not so much on the unstructured data. “We define a knowledge product with four elements – sensitivity, unstructured data quality, metadata for humans, and metadata for AI tools.”

"One key difference to note about them is that in the structured data world, data products are typically consumed by analytics, data and AI teams. Knowledge products, conversely, I think will be consumed by everyone."

Overall, Collibra’s system offers a multitude of solutions, including AI-based taxonomies which can automatically create meaningful taxonomies and segments directly from the data. This dramatically cuts down on the lengthy effort that manual mapping requires from subject matter experts.

This shift lets companies focus on areas where effort brings real value. For example, businesses can scan supplier contracts to identify auto-renewal clauses that trigger in 30 days. Enterprises can prioritise high-value, validated items that need human review. As technology evolves, this entire system is moving toward AI-managed workflows, representing a significant advance toward an autonomous enterprise.

• 85 percent of enterprise content is unstructured (documents, transcripts)
• Unstructured data is the new foundation for scalable AI
• Governance must be unified (structured + unstructured) to simplify tools and serve AI agents effectively.
• Manual data labeling is impossible. 
• AI/LLMs must automate metadata generation via a Continuous Tagging System.
• Leaders should adopt the Knowledge Product idea—a governed, AI-ready asset for unstructured data consumed by the entire enterprise.

0:00: Intro: The Exciting Moment in Tech

1:42: Unified Governance via Dz Acquisition

4:27: The 4-Layer Unstructured Data Funnel

7:35: Automation: Continuous AI Tagging

12:55: The Value of Knowledge Products

Collibra frees your data from the constraints of silos by unifying data and AI governance across your entire ecosystem, regardless of source or compute engine, for ultimate flexibility in how you manage data. Our Collibra Platform gives you automated visibility, control and tracing from input through output, and it automates documentation and data traceability for AI use cases to power speed, data observability and safety. Our enterprise metadata graph enriches data context with every use, and our intuitive UX brings technical and business users into the fold to access and steward data.