A year after predicting GenAI's sustainability crisis, the 2024/25 financial data tells a concerning story. OpenAI's $10 billion revenue comes with $5 billion losses, whilst Anthropic burns $3-4 billion annually. With the sector consuming over $100 billion in venture funding and Big Tech spending $250 billion on AI infrastructure, we're witnessing what some critics are now calling a "subprime AI crisis" – an entire industry built on services sold at massive losses. The transparency initiatives from Hugging Face's AI Energy Score and the shift towards open source models (now 46% of enterprise preference) offer hope for a more sustainable GenAI 2.0 focused on efficiency over scale.

A roundup of themes we're seeing in Private Equity technology, and some recommended resources.

In this episode of Beyond the Hype, Oliver Cronk is joined by Ani Allen from OutSystems and Ryan Grey from Marra to explore the spectrum from No-Code, through Low-Code, to High- (or Pro-?) Code, and consider how AI is reshaping the development experience.

Last year, I received a combined autism and ADHD (AuDHD) diagnosis. It was a pivotal moment that helped me reframe my life, and I decided early on to be open about my diagnosis, both to advocate for myself and to let others know they’re not alone.

Data enables faster and more accurate due diligence, informs operational transformation post-acquisition, and supports more effective positioning when it comes time to exit. This post outlines the role of data across each of these key stages.

If you’re working on an entirely new product with no dependencies on any existing system, you’re free to decide how best to build, test and deploy the product. However, if you’re modernising a legacy system, the approach to development and testing is less straightforward. In this blog post, I look at some of the key considerations that should inform the test strategy for a modernisation project. I also explain why it’s important to go beyond the hype and use the right tool for the right job at the right time.

On the route to delivering and realising value, the aim is to reduce the number of twists and turns in the software development pipework and improve flow. In this blog post, I’ll share some insights into practices and processes that reduce flow and what you can do to improve flow. It’s important to point out this is purposefully methodology-agnostic; I want to get beyond the rituals and practices of the different approaches and focus instead on the underlying best practices.

Pride Month is both a joyful celebration and a powerful protest — a moment to honour how far we’ve come, and to spotlight how far we still have to go. But every year, as the corporate logos change their colours and the parades grow ever larger, the same question resurfaces: Do we still need Pride? The answer is a resounding yes. In this post, I explain why.

What? You want to test in production? You must be joking!

In this post, I explore what a Data Lakehouse is, how it works, and whether it delivers on its promises—covering core features, formats, real-world patterns, and platform realities.

We’ve been working with a market-leading energy trading firm that’s taking first-mover advantage in the energy sector by adopting a desktop interoperability platform. In this blog post, I’ll provide an overview of how they work, outline the main benefits they deliver, and flag some trade-offs to bear in mind.

In this episode, I dive into the world of Event-Driven Architecture (EDA) with Tom Fairbairn from Solace and Scott Logic’s Gordon Campbell. The discussion explores whether EDA has matured beyond the hype into a practical strategy for modern systems integration, or if it’s just another architectural buzzword.

From our experiences and chats with clients on deploying AI, we can perhaps identify what looks like a spectrum of architectural approaches for enterprise AI deployment. In this post, I explore this spectrum with the goal of helping you consider where your organisation is today and where it might best position itself.

The public sector faces growing challenges — tight budgets, rapid tech change, rising expectations, and staff shortages — often leading to decision paralysis or rushed mistakes. Many agencies found a solution by partnering with private consultancies to provide agile and robust solutions. True success relies on combining public sector knowledge with private sector expertise, showing that collaboration is key to stronger, more resilient communities.

We have been helping a client migrate their trading platform to a new version of a third-party API. The migration is more interesting than usual for a number of reasons, so I thought it might be useful to share some insights from our experience.

Large Language Models are powerful tools that can greatly enhance software developers' productivity, but for junior developers starting a career in tech, they may hinder long-term growth by abstracting away essential programming fundamentals.

The bench is a key part of the project-based nature of consulting and managing time on bench effectively is a key skill for consultants. Here are some things every consultant should reflect on between engagements.

Features or Foundations. Where do you start. What are the pros and cons of building fast or building the blocks to build on.

Despite the UK being the world's third-largest AI industry, only 1 in 6 UK firms are actively using AI due to financial, skills, and risk concerns, but businesses can bridge this "valley of doubt" by adopting measured "AI in the loop" approaches with human oversight, right-sized models, and government frameworks like Bridge AI to minimise business, societal, and environmental risks.

In this post I'll discuss ways of estimating the emissions caused by your Cloud workloads as a first step towards reaching your organisation's Net Zero goals.