Brendan Foxen

£10Bn [ringfenced] tech fund for NHS to help drive reform through digital / AI innovation. Of course, that’s good news for the NHS and proven suppliers to it, like workpal However, whilst funding is vital, the challenge in releasing that in a way that drives true reform lies in the ambition and leadership of those in the system now charged with spending it to guarantee a return on that investment. And, let’s remember, the espoused vision to empower citizens to own their health should be true also for ALL staff to be equally engaged to own and drive their performance, all in line with the 10YP and subsequent more detailed strategies. It’s all about change and people / leaders appetite to do things differently. Our wonderful NHS workforce are more than ready, as our work with a variety of Trusts and Staff up & down the country shows. The performance framework and incentives should help drive meaningful action. Interesting couple of years ahead. #nhs #nhsengland #nhsproviders #nhsprofessionals #nhsemployers #digital #AI #workforce #performance #culture #productivity HPMA Healthcare People Management Association

8

1 Comment

Why we built the Pro-Test AI Hub specifically for the "Regulated Era" of Engineering. As we scale Pro-Test Technologies in the UK, the most common question I get from CTOs is: "How do we use AI to build faster without losing our grip on Governance?" The answer isn't "more testers." It’s Smarter Guardrails. Today, I’m pulling back the curtain on the Pro-Test AI Hub—our cloud-based engine designed to turn "AI-Speed" into "Regulatory Confidence." The Hub isn't just a testing tool; it’s a Quality Orchestrator: - The DORA Mapper: It automatically maps every test execution to specific ICT Risk pillars, giving you a real-time "Compliance Scoreboard" for the FCA. - Deterministic Agent-Testing: Our agents don't just "guess." They use property-based logic to stress-test AI-generated code, catching the "Logic Ghosts" that standard QA misses. - The Resilience Dashboard: A single pane of glass for UK leadership to see exactly where their "Impact Tolerance" stands before a release goes live. We didn't build this to replace engineers. We built it to empower them to ship at Bengaluru-velocity with London-standard safety. UK Engineering Leaders: Are you still managing quality in spreadsheets, or are you ready for a real-time Resilience Control Tower? #AIHub #ProTestTechnologies #UKFintech #DORA #QualityEngineering #SaaSLaunch #LondonTech #OperationalResilience

5

I thoroughly enjoyed chatting with Anita Allen (OutSystems) and Oliver Cronk (Scott Logic) on “Beyond the Hype.” where we unpacked how AI is shaping the future of low-code We dove into why “citizen development” is more nuanced than the hype suggests, and shared practical low-code starting points. Insights that matter for Marra clients and anyone building using low-code or AI dev tools. 🎧 Full episode in comments. #LowCode #AI #CitizenDeveloper

21

6 Comments

𝗚𝗼𝗼𝗴𝗹𝗲 𝗶𝗻𝘁𝗿𝗼𝗱𝘂𝗰𝗲𝗱 𝘁𝗵𝗲 𝗻𝗲𝘄 𝗚𝗲𝗺𝗶𝗻𝗶 𝗜𝗻𝘁𝗲𝗿𝗮𝗰𝘁𝗶𝗼𝗻𝘀 𝗔𝗣𝗜. It’s a complete rethought of how we interface with LLM APIs. For the past few years, the dominant pattern for interacting with large language models has been: text in, text out.This is a stateless exchange where the client holds all context and the model regurgitates responses. That served its purpose. But it’s no longer enough for the next chapter of AI. We need stateful, agentic systems and not just clever autocomplete. The Gemini Interactions API is a meaningful step in that direction. It provides a unified interface for models and agents, enabling developers to build systems where context, history, and long-running tasks are first-class citizens. That means: 🔹 Server-side state handling helps in conversations and tasks persist without shipping massive histories back and forth. 🔹 Background execution lets you kick off complex workflows and check status later, elevating what we can reliably automate. 🔹 Built-in agent access like Deep Research, Code Executor provide powerful, multimodal reasoning out of the box. This is less about hyped “assistant agents” and more about structuring computation and cognition in a way that scales. When you start thinking about AI as a platform for interaction patterns and not just prompt/response , the API’s design becomes compelling. Tools like this redefine how developers can build with AI today. They shift the focus from brittle client-side state management to robust server-driven flows, which I think will fundamentally change how we compose AI into real systems. If you’re building workflows that rely on persistent context, multimodal reasoning, or orchestrating agents over time, this API is worth exploring. Check out how at Intugle we make AI agents future proof and ensure you are able to use them on your data at scale. 🚀 #AI #GenerativeAI #GeminiAPI

21

Department for Science, Innovation and Technology Exceptional initiative by the UK Sovereign AI Unit. To build truly sovereign autonomous labs, the UK will need a reasoning layer that embeds peace- and nature-aligned governance at the core. NAPAI TLP (Nature-Aligned Peace-Centric AI Trust Logic Protocol), supported by SCP-1 (Sovereign Compliance Protocol-1), is designed exactly for this purpose — ensuring AI-driven experimentation remains safe, transparent, and aligned with long-term national scientific interests. We are pleased to share that we have submitted a proposal for a hybrid Wet-Lab + Digital Twin Autonomous Laboratory governed by NAPAI TLP. The vision is simple: an autonomous experimentation platform that is sovereign, safe, ethical, and capable of accelerating scientific discovery without relying on foreign reasoning architectures. Our conceptual architecture includes: • A peace- and nature-aligned sovereign reasoning layer • SCP-1 for transparent, auditable autonomous decisions • Wet-lab robotics integration (Automata, Opentrons, Synthace) • AI-driven hypothesis generation and closed-loop optimisation • Digital twin simulation for rapid and safe iterative discovery • A future pathway for humanoid robotics such as Tesla Optimus Autonomous labs will shape the future of science. We believe the UK can lead if sovereign logic frameworks are embedded from the beginning. Glad to contribute to this national effort. #SovereignAI #AutonomousLabs #AIEthics #ScientificDiscovery The Alan Turing Institute

2

We're continuing to move down a path of integrating Model Context Protocol (MCP) to both orchestrate some of our work with AI at BITE Data, and also allow external users to access our data with their own AI agents. Its been interesting architecting the solution - as the article here from Anthropic states, although many of the problems using the protocol feel novel, including context management, tool composition, state persistence - they have known solutions in the software engineering / architecture world. The line of sight between API endpoints, the MCP SDK, data tools that have to be built, server prep, network rules and other components in the implementation architecture are still the same. Its just then when you finally connect and run jobs, its a bit like how Simba felt when Timon and Pumba showed him their hidden paradise in Lion King 1. This for sure is not something you can just pick and do without engineering knowledge and the ability to build loosely coupled, distributed software architectures that integrate fairly complex data pipelines and stores. The consumption of tokens piece is important as well, as the article notes. https://lnkd.in/eR3cRaUc

16

🧩 New release at Pigment: Granular Deployment is here! We’ve gone from deploying entire Workspaces, to Applications, and now all the way down to individual changes within those Applications. Smaller changes, more control, clearer visibility into what’s actually being deployed — and a much better experience when multiple people are working in the same Application at once. Really, really fun one to see go live 🎉 🔗 Here's all about it: https://okt.to/hZMO1H

8

Just 161 companies in the UK currently provide AI Assurance services (out of a total of about 5 million companies). We’re now making that 162 with ADVANTAGE AI (give or take a bit). Using AI tools can seem simple. Implementing AI in the workplace can be a much more complex undertaking. The loss of customer confidence and trust combined with regulator penalties for AI failings can be significant. And let’s not forget the personal accountability of executives and senior leaders for those in FCA regulated companies. The independent 360-degree AI assurance assessments by ADVANTAGE AI can provide your management team with peace of mind with clarity on your organisations strengths and risks and a bespoke roadmap of prioritised remediation actions. — “In November 2024, Department for Science, Innovation and Technology (DSIT) published its first analysis of the state of the UK’s AI assurance market. We identified an estimated 161 UK-based assurance firms…” Source https://lnkd.in/eei2p28c #ai #artificialintelligence #assurance #risk #governance #regulator #compliance Department for Science, Innovation and Technology Financial Conduct Authority

4

What is the AIOS (AI Operating System) by Implement AI ? On Friday in London at our event ‘Your Future Digital Workforce’ we took business owners from across the UK & Ireland through the three layers of our AIOS platform, as well as real world use cases and applications. Aalok Yashwant Shukla and I explain it in more detail and how the powerful combination of multichannel and multimodal AI agent teams, our Agentic CRM and our Agentic Task Engine, can help drive growth in your business. Grow you workforce. Not your payroll. Implement AI #AI

23

5 Comments

🔥 𝗨𝗣𝗗𝗔𝗧𝗘: Synthesia CEO Riparbelli on Tech Opportunities in Europe Synthesia CEO Victor Riparbelli says ongoing political tension has created better opportunities for entrepreneurs to build companies as focus shifts away from regulation and into the technology and defense sectors. He speaks with Bloomberg's Tom Mackenzie at the Founders Forum in Oxfordshire.

1

When the UK AI Opportunities Action Plan was announced, the wording around the AI centres for excellence suggested that investment and interest would concentrate around the south. I hoped that this would not be the case, but it seems that once again, the north is overlooked. https://lnkd.in/eYvCMnrK

15

2 Comments

Very proud and privileged to be advising this Leadership team to help them reach their AI ambition. It is a bold vision we are working too. Truly empowering their people, alongside an AI platform, so they can deliver EVEN MORE outstanding services to UK citizens. Reinvesting the efficiency gains we make to deliver human centric services that Government and communities really care about. Its a transformation that is powerful, challenging and exciting! Kenny Morris Marco Jetmir Ndrecaj Neil Gladstone Adrian Fieldhouse Teresa Fox Jonti Dalal-Small Thomas Gladwin Rebecca Moseley Caitlin Clark Fiz Yazdi Gary Craven ChMC #hybridintelligence #soprasteria #soprasterianext #responsibleai

24

1 Comment

Check out our blog to see how businesses can succeed at getting AI beyond PoCs and into production. I can say through experience that getting something to work 80% of the time with AI is easy, basically hackathon easy, but getting something ready for the real world is another challenge altogether.

9

3 Comments

I read the UK government’s announcement to spend £100M buying emerging chip technology from British companies with interest (https://lnkd.in/eXmgn4GR). In order to be truly useful, we have to see this investment leveraged to bring in significantly more private capital to help our excellent semiconductor start-ups scale. The FT report highlights the challenge: US private investment in AI was $109.1Bn in 2024 vs $4.5Bn in the UK. We clearly have some way to go. To really bridge this gap and achieve meaningful scale, we not only need government backing, but we have to produce truly investible businesses which match great technology with strong market positioning. This is all about the product-led approach about which I am passionate. If you're a CEO, VC, or PE investor and want to talk about how product-led thinking and optimized system design can fundamentally transform your scaling business, reach out to me on LinkedIn or at https://lnkd.in/ehShMyu9 #DeepTech #VentureCapital #SemiconductorStrategy #ScaleUp #ProductLedGrowth

43

4 Comments

I left Palantir Technologies to build Calibre. The reason traces back to working on the FDP Marketplace at the NHS, where partner companies shipped apps on top of the platform. Teams would build something that worked - a clinical decision support tool that could shorten triage, flag a missed diagnosis, save a life. They'd put every control in place, stress test the system. Then disappear for 12-18 months into EU MDR before a clinician could touch it. 12-18 months between "safe enough to use" and "actually being used". Because the system that exists to verify safety hasn't been rebuilt for the volume or the velocity of software/hardware being pushed through it. AI agents are about to multiply the amount of evidence in front of an auditor by an order of magnitude. Verification becomes the rate limit on everything. That's why we started Calibre.

129

1 Comment

We often assess whether our systems are robust enough for scale and forget that resilience isn’t only a system question but also a governance one. Control and accountability at the data layer are vital. The techUK article Robust AI Governance as a Path to Organisational Resilience gets right to the point: governance is a business-critical function that frames how organisations manage risk, comply with emerging AI and data regulations, and ensure operational continuity. For CISOs and their peers, the key question is shifting from “Is the data secured?” to “Can we prove how it’s being accessed, by whom, and why, across all environments, including AI pipelines?” This is where Self-Sovereign Data™ becomes foundational. By keeping data at source (zero-copy), embedding access policy logic directly into the data asset, and enforcing access through portable permissions, governance becomes continuous, tamper-proof, and scalable across jurisdictions. It’s a model built for the governance mandates of 2025 and beyond: continuous, provable control, even outside the perimeter. #SelfSovereignData #datagovernance #datasecurity #confidios

11

1 Comment

If AI feels exciting but confusing then the National AI Centre has something to help! The "Guidance for AI Adoption: Foundations" published earlier this week by National AI Centre makes it easier than ever for businesses to get started with AI. Heres the secret sauce: 1️⃣ Decide who’s in charge: Pick someone to own AI in your business and write a simple policy. 2️⃣ Understand your impact: Think about who your AI might affect, what could go wrong, and how people can speak up if it does. 3️⃣ Manage risks: Spot issues early, track what happens, and fix things fast. 4️⃣ Be open: Tell people when they’re interacting with AI and keep a list of where you use it. 5️⃣ Test and check: Try it before you trust it. Keep an eye on it over time. 6️⃣ Keep humans in control: Always make sure a person can step in, pause, or switch it off if needed. Step 5 (Testing) is probably one of the hardest steps. You never know what someones going to throw into the machine and what it will spit out. Don't skimp on TESTING! If you get the steps above right (its not always as easy as it sounds), then you'll be punching above your weight. You'll build trust, accountability, and resilience as AI becomes part of daily operations. When you’re ready to go deeper, there’s a follow-up guide too! It gives step-by-step instructions for bigger or riskier projects. Innovation and transformation is still hard work, but it’s never been this accessible.

2

The first UK-US tech agreement announced today claims to have the potential to create more than 5,000 jobs based on more than £31 billion investment into the UK AI and tech infrastructure, including Microsoft’s largest ever commitment to the UK. https://lnkd.in/e7hPEwDd #UK #US #Trade #ClimateTech #AI #Investment #DrugDiscovery #SpecialRelationship #NewPaloAlto

8