Amy Boyd

No need for backend coding any more due to AI? I took my son to a sixth form college last night - he's interested in a Computer Science A-level (for those outside the UK, this is for ages 16-18 as a preparation for university). The college offered a new course in Computing, but it was focused on application development, project planning, front end stuff etc. Apparently this is an initiative from the UK government's Department of Education, the reason being "because no-one will need to work on backend code, that will all be done by AI". I found this very odd and a little worrying. I'm sure the DofE and government generally is deluged by those enthusing about AI - but is it really the case that no-one will need to know how the backend works? Personally I've always thought that a knowledge of the entire stack is important, and even with AI surely we're going to still need people to build low level code, invent new languages and compilers, tune for performance etc. In my experience this is where true creativity is required, to make computers do things they weren't expected to be able to do.

72

30 Comments

In our latest Tech Talk, part of our series highlighting some of the brilliant engineers across Tech Catalyst, our innovation business unit, I’m joined by Antonia Berry, one of our lead data engineers and architects, to dive into a fundamental — and often underestimated — pillar of successful AI: data quality. We unpack why high-performing AI begins with high-quality data, and explore emerging capabilities inside MS fabric & data virtualisation, which many of our clients use as a common data layer behind multiple AI models and providers. Stay tuned for the next instalment as we go even deeper into the world of #AI development. #ArtificialIntelligence #AI #GenAI #DataQuality #DataEngineering #DataArchitecture #MicrosoftFabric #DataVirtualisation #TechTalk #TechCatalyst #DigitalTransformation #CloudInnovation #Analytics #FutureOfTech #PwC Umang Paw Peggy Gondo Antonia Berry Euan Cameron Dominic Mac Rob McCargow Stephen Mills

64

3 Comments

I know that this is not a direct comparison but this image points out the silicon binge most of our applications rely on to just about operate. It is not just because of feature bloat. The main issue is lack of appreciation of profiling and optimisation. This is largely driven by two key myths: 1. Developer time is more expensive than computer hardware 2. Premature optimisation is evil Let's assume that your application requires one server (any size) to run on and you compare the cost of that server to a developer's annual salary, yes, then a developer's time may seem much more expensive. But you would only make such comparisons if a developer is to spend a whole year to optimise an application to save on hardware. In most cases, it will be a few weeks worth of an effort, if not days. The cost of developer time in optimising software will be paid off fairly quickly through the hardware savings. Secondly, it is speculative optimisation that is evil, not knowing your performance and capacity requirements and trying to make your software "as fast as possible". That can also lead to regressive behaviours like buying hardware that you don't need and making your implementation too customised or complex. In a nutshell: 1. Know your cross/non functional requirements 2. Optimise your code as you write it 3. Keep software simple, modular, decoupled and scalable 4. Profile early and often to discover and resolve performance hotspot This will keep your infra and cloud costs in check #technology #strategy #leadership #softwareengineering #optimisation

1,257

123 Comments

With all the talk and learning around GenAI... can I suggest some great fundamentals.. Modern Software Engineering by Dave Farley and understanding your codebase, not just current static state, but the social behaviours over time - Your Code as a Crime Scene by Adam Tornhill.

135

7 Comments

I got this email a couple of days ago, and that's when I realised that what started as system designs written on paper has become a platform that can genuinely make a difference in apprenticeship applications. Looking back to when I was applying last year, I realised that I had no idea what I was doing, and no clear guidance on how to stand out among thousands of applicants. That’s why I built Proofbase, and being able to combine my experience with both apprenticeships and engineering to build a geniunely useful tool has been super rewarding. Students can build professional portfolio websites and CVs, find and track their applications, and get AI feedback to improve, all specifically designed for apprenticeships. If you’re applying this year or know someone who is, I’d really recommend having a look at it, a lot of time and knowledge has been put into this, and it's already been endorsed by multiple other apprentices at top companies. 👇 https://proofbase.app

73

👏Building the UK’s next generation of AI talent starts with applied skills! QA Ltd has teamed up with NVIDIA as an official NVIDIA training provider, launching the first NVIDIA-certified and QA-delivered AI apprenticeship programme designed to build real, job-ready capability inside organisations. By combining QA’s market-leading AI training with NVIDIA certification via the Deep Learning Institute, this  collaboration gives UK employers a practical, levy-funded route to develop AI skills at scale. �� Read the full announcement here: 🔗 https://shorturl.at/MwpEJ 🔗 #AI #Apprenticeships #DigitalSkills #AIEducation #FutureOfWork #QAxNVIDIA #NationalApprenticeshipWeek

112

4 Comments

Who's really driving the UK's AI revolution? 🤫 Forget what you know about tech hubs and 10x software engineers. Now, nurses, librarians and surveyors are emerging as the next generation of AI talent. Our latest research reveals an "invisible AI workforce" that's powering adoption behind the scenes. We're talking about a diverse group of professionals, aged 19 to 71, who are embracing it in ways you wouldn't expect. ↳ 67% of people applying AI at work are in non-tech roles. ↳ From healthcare to construction and local government - it’s not just digital industries that are embracing the tech. ↳ AI talent hotspots are emerging outside of London, in places like Trafford, Birmingham, and Cheshire West and Chester. This is a massive opportunity for organisations to unlock hidden potential and drive innovation from within. Meet the UK’s #InvisibleAIWorkforce in our latest research: https://lnkd.in/dp5mB6Y8

72

6 Comments

National Apprenticeship Week is over. So, what happens now? A week later, with space to reflect, it’s clear that the real work begins once the noise dies down. NAW creates momentum, but it’s what we all choose to do next that shapes the system apprentices experience every day. This year, the message came through clearly at every event I attended. Apprenticeship reform is challenging, but the system is not broken. It's just changing. And the biggest improvements come from listening to apprentices early, listening often, and acting on what they say. The Association of Apprentices' Insight Report 2026 shows this. Apprentices know what is working and they know what isn't. They are not asking for perfection. They are asking for clarity, fairness, smoother onboarding and support that meets them where they are. So, what comes next? Here is what I would love to see: 👉 Apprentice voice built into policy reform conversations from the start 👉 Cross sector communities staying connected beyond NAW 👉 Employers and providers asking apprentices what they need and listening 👉 Celebration and recognition becoming part of everyday culture rather than something saved for one week a year The system works best when apprentices, employers, providers and policymakers all work together all year round, not just once a year. For me, NAW 2026 has shown that apprenticeships thrive when people feel seen, supported and included. And if we keep listening to apprentices, we can build a system that becomes better every year. Apprenticeships can change lives. We have a responsibility to do them right. #NAW2026 #Apprenticeships 

56