Introducing the Databricks data lakehouse architecture as a modern solution for data management. It explains how this architecture combines the flexibility and cost-effectiveness of data lakes with the analytical capabilities of data warehouses, aiming to eliminate data silos and accelerate data and AI initiatives. The book discusses the limitations of traditional data warehouses and data lakes, highlighting the need for a unified platform that supports diverse data types and workloads, including business intelligence (BI), machine learning (ML), and artificial intelligence (AI). Furthermore, the text explores the underlying technologies like Delta Lake and Unity Catalog, which provide data reliability and unified governance, and details how the data lakehouse forms the foundation for data intelligence platforms. Get the book for free: https://www.databricks.com/resources/ebook/the-data-lakehouse-for-dummies?scid=701Vp00000MdUO9IAN&utm_source=databricks&utm_medium=email

The European Union's Digital Services Act (DSA) and Digital Markets Act (DMA), a legislative package designed to establish a safer and more equitable online environment. The DSA focuses on protecting online users' fundamental rights and addressing the spread of illegal content and disinformation by imposing tiered obligations on digital services, with the strictest rules applying to Very Large Online Platforms (VLOPs) and Very Large Online Search Engines (VLOSEs) based on user numbers. It aims to increase transparency in content moderation, prohibit targeted advertising based on sensitive data or for minors, and provide users with mechanisms to report illegal content and challenge moderation decisions. The DMA targets large "gatekeeper" platforms to ensure fair business practices and foster competition and consumer choice within the digital market. Both acts include significant penalties for non-compliance and are enforced through a cooperative framework involving the European Commission and national Digital Services Coordinators.

In this TechSence podcast we look at a new publication from Eurostat.

This publication is an interactive online report from Eurostat, which provides statistical data and insights into the state of digitalisation across Europe as of 2025. It examines the digital skills of individuals, noting that a significant portion still lack basic proficiency despite high internet usage. The publication also analyses how businesses adopt technology, including the use of artificial intelligence and cloud services, and highlights the progress and challenges faced by enterprises in reaching digital intensity targets. Furthermore, it explores people's online activities, such as e-government service utilisation, internet banking, and online shopping trends, while also touching upon the Internet of Things and the recycling of digital equipment. The data presented is based on annual surveys focusing on ICT use in households and enterprises.

Source: Digitalisation in Europe – 2025 edition - Interactive publications - Eurostat

Uphill struggle ahead for Govt’s use of AI as PAC report reveals the scale of the challenge - Committees - UK Parliament

The UK Public Accounts Committee (PAC) has identified significant challenges in the government's efforts to adopt artificial intelligence (AI) within the public sector. A primary concern is the prevalence of outdated technology and poor-quality data, which hinder AI implementation. In 2024, approximately 28% of central government IT systems were classified as 'legacy' systems—outdated and unsupported—posing obstacles to AI integration.

The PAC also highlights a lack of transparency in the government's use of AI, which could erode public trust. By January 2025, only a limited number of records had been published on a government website dedicated to algorithm-assisted decision-making transparency. Addressing public concerns about the sharing of sensitive data in AI applications is deemed essential.

Furthermore, the dominance of a few large technology suppliers in the AI market raises concerns about competition and innovation. The government's procurement approach may lead to over-reliance on specific companies, limiting adaptability.

Persistent digital skills shortages present another barrier. In 2024, around half of civil service digital and data roles remained unfilled, with 70% of departments reporting difficulties in recruiting and retaining staff with AI expertise. The PAC questions whether planned digital reforms will effectively address this issue.

To drive change, the PAC recommends embedding senior digital officers within departmental leadership teams. Sir Geoffrey Clifton-Brown MP, Chair of the Committee, emphasizes the need for transparency to build public trust and ensure the safe and ethical adoption of AI in government services.

Many companies are heavily investing in artificial intelligence (AI), yet they often lack the necessary infrastructure and skilled personnel to fully realise its potential. A significant challenge is building a robust data infrastructure, with many firms addressing data issues inconsistently on a project-by-project basis. Poor data quality, stemming from messy and siloed information, hinders even advanced AI models. Furthermore, organisations struggle to find talent in areas like machine learning and face internal resistance to AI adoption, sometimes leading to waning enthusiasm. To succeed with AI, companies need to develop cohesive data strategies, invest in automation for data management, upskill their workforce, and foster a culture that embraces AI.

This discussion is based on the following CNBC article:

AI implementation projects are far from intelligent inside companies

https://www.cnbc.com/2025/01/22/ai-implementation-isnt-being-done-intelligently-inside-most-companies.html

We introduce the Office of the EU Data Protector and discuss one of their recent reports on Deepfakes and how what the EDPS is doing about them.

Sources:

Podcast based on article: Machine unlearning | European Data Protection Supervisor

https://www.edps.europa.eu/data-protection/technology-monitoring/techsonar/machine-unlearning_en

TechSonar | European Data Protection Supervisor

https://www.edps.europa.eu/data-protection/technology-monitoring/techsonar_en

We discuss a useful article on LinkedIn where we find out what big data sets are available for research and study:

This text details numerous publicly accessible datasets. Several websites are highlighted, including "Our World in Data" focusing on global issues, "Get Open Data" offering UK-centric data, and "Open Data Northern Ireland" providing government and public sector information. Specific datasets are listed, covering areas like health, economics, and crime. Additional resources are also included, such as links to repositories containing New York City taxi data and articles compiling various big data sources. The overall aim is to provide a comprehensive overview of readily available data for research and analysis.

url: https://www.linkedin.com/pulse/available-datasets-study-big-data-science-research-les-black-llbzf

Microsoft details three proven patterns for using their Azure OpenAI Service to build AI-powered applications. Pattern 1 shows how to create a GenAI chatbot, Pattern 2 explains building a personalisation engine, and Pattern 3 guides users through creating a custom copilot. We discuss Azure OpenAI's capabilities, including its various models (like GPT-4), multimodal features, and enterprise-grade security. Real-world customer success stories illustrate the benefits of using Azure OpenAI for increased productivity and improved customer experiences.

htts://portal.azure.com