https://ives-yim.org.uk/writing/ Articles on AI in engineering companies, technology leadership, strategy and connected products. en-gb https://ives-yim.org.uk/writing/agents-at-work-in-engineering/ https://ives-yim.org.uk/writing/agents-at-work-in-engineering/ Sun, 27 Sep 2026 09:00:00 +0000 The first articles in this series were about what agents are and when not to use them. This one is about where they are genuinely useful in an engineering company. https://ives-yim.org.uk/writing/agents-with-keys/ https://ives-yim.org.uk/writing/agents-with-keys/ Sun, 27 Sep 2026 09:00:00 +0000 An AI assistant that answers questions has a limited capacity for harm. An agent with access to your systems is different. It holds keys: credentials, permissions and tools that let it read and change things on the organisation's behalf. https://ives-yim.org.uk/writing/ai-drafts-engineers-decide/ https://ives-yim.org.uk/writing/ai-drafts-engineers-decide/ Sun, 27 Sep 2026 09:00:00 +0000 Ask a senior engineer how much of their week goes on documentation and you will rarely get a small number. Root-cause reports, service bulletins, change impact assessments, failure analysis updates, test plans, compliance files, responses to customers and regulators. All of it necessary. https://ives-yim.org.uk/writing/ai-is-not-my-first-rodeo/ https://ives-yim.org.uk/writing/ai-is-not-my-first-rodeo/ Sun, 27 Sep 2026 09:00:00 +0000 What seven hype cycles taught me about this one https://ives-yim.org.uk/writing/building-software-with-coding-agents/ https://ives-yim.org.uk/writing/building-software-with-coding-agents/ Sun, 27 Sep 2026 09:00:00 +0000 I have written before about leading the build of a production platform with AI, and about how it removed the translation layer between an architect's thinking and the code that gets written. That article was about what changed. https://ives-yim.org.uk/writing/closing-the-loop/ https://ives-yim.org.uk/writing/closing-the-loop/ Sun, 27 Sep 2026 09:00:00 +0000 Early in my career, I was asked to look at a service department that was quietly draining hundreds of thousands a year from a company's margin. The information that explained it was all there, in job sheets and spares records, and nobody had put it together. https://ives-yim.org.uk/writing/coding-agents-need-experienced-engineers/ https://ives-yim.org.uk/writing/coding-agents-need-experienced-engineers/ Sun, 27 Sep 2026 09:00:00 +0000 A coding agent will write code for anyone. That is both its great promise and its great danger. In the hands of an experienced engineer, it is the most productive tool I have used in my career. https://ives-yim.org.uk/writing/coding-agents-strengths-and-weaknesses/ https://ives-yim.org.uk/writing/coding-agents-strengths-and-weaknesses/ Sun, 27 Sep 2026 09:00:00 +0000 Most of what is written about coding agents is either breathless or dismissive. They will replace developers, or they produce rubbish. Neither matches my experience. I use them daily, and have used them to build a production platform and to develop an AI system of my own. https://ives-yim.org.uk/writing/connected-products-vs-ai-attackers/ https://ives-yim.org.uk/writing/connected-products-vs-ai-attackers/ Sun, 27 Sep 2026 09:00:00 +0000 Most of this security series has been about protecting the AI systems an engineering company uses. This article turns to the products it makes. Connected products, from vehicles and wearables to building controls and industrial equipment, have always been targets. https://ives-yim.org.uk/writing/cost-and-infrastructure-of-agents/ https://ives-yim.org.uk/writing/cost-and-infrastructure-of-agents/ Sun, 27 Sep 2026 09:00:00 +0000 The first time an organisation runs agents seriously, the bill is often a surprise. A task that cost pennies as a single question can cost many times more when an agent works through it. https://ives-yim.org.uk/writing/defending-with-ai-and-testing-your-ai/ https://ives-yim.org.uk/writing/defending-with-ai-and-testing-your-ai/ Sun, 27 Sep 2026 09:00:00 +0000 This is the last article in the security series, and it draws together the three series so far on AI in engineering companies. The first described five levels of AI value, from finding information to closing the loop from field data. https://ives-yim.org.uk/writing/five-levels-of-ai-value/ https://ives-yim.org.uk/writing/five-levels-of-ai-value/ Sun, 27 Sep 2026 09:00:00 +0000 Most of the conversations I have about AI in engineering companies start and end in the same place: "Could we use it to search our documents?" It is a good question, and the answer is yes. I have written about it separately. https://ives-yim.org.uk/writing/from-documents-to-data/ https://ives-yim.org.uk/writing/from-documents-to-data/ Sun, 27 Sep 2026 09:00:00 +0000 In the first article in this series I described five levels of AI value in an engineering company. The second level, structure, is the one I most often find missing. https://ives-yim.org.uk/writing/from-the-manual-to-the-product/ https://ives-yim.org.uk/writing/from-the-manual-to-the-product/ Sun, 27 Sep 2026 09:00:00 +0000 The first two levels of AI value in an engineering company work with what the organisation already has on file: its documents, and the structured facts extracted from them. The third level is different in kind. https://ives-yim.org.uk/writing/guardrails-autonomy-is-earned/ https://ives-yim.org.uk/writing/guardrails-autonomy-is-earned/ Sun, 27 Sep 2026 09:00:00 +0000 Every conversation about AI agents eventually reaches the same question: how much should we let it do on its own? https://ives-yim.org.uk/writing/how-data-leaks-out-of-ai-systems/ https://ives-yim.org.uk/writing/how-data-leaks-out-of-ai-systems/ Sun, 27 Sep 2026 09:00:00 +0000 When organisations worry about AI and data, they usually worry about one thing: sending confidential information to an AI provider. That is a real concern, and I discussed it in Where should your AI live? . https://ives-yim.org.uk/writing/keeping-your-code-local-and-private/ https://ives-yim.org.uk/writing/keeping-your-code-local-and-private/ Sun, 27 Sep 2026 09:00:00 +0000 For many engineering companies, source code is the most valuable thing they own. Proprietary firmware, control algorithms, years of accumulated product knowledge expressed in software. It is also exactly what a coding agent needs to read in order to be useful. https://ives-yim.org.uk/writing/local-cloud-and-frontier/ https://ives-yim.org.uk/writing/local-cloud-and-frontier/ Sun, 27 Sep 2026 09:00:00 +0000 Most engineers using coding agents use one model for everything, usually the most capable one they have access to. It is the simplest approach, and for an individual it is often reasonable. https://ives-yim.org.uk/writing/managing-a-token-budget/ https://ives-yim.org.uk/writing/managing-a-token-budget/ Sun, 27 Sep 2026 09:00:00 +0000 In What tokens are, and why they matter , I explained that tokens are the unit in which AI does its work, and that coding agents consume them in large quantities, mostly by rereading context at every step. https://ives-yim.org.uk/writing/running-agents-in-production/ https://ives-yim.org.uk/writing/running-agents-in-production/ Sun, 27 Sep 2026 09:00:00 +0000 Getting an agent to work in a demonstration is easy. Getting it to work reliably, every day, for months, while the data changes, the models change and the people using it change, is an engineering discipline of its own. It is also where most agent projects quietly fail. https://ives-yim.org.uk/writing/security-of-ai-written-code/ https://ives-yim.org.uk/writing/security-of-ai-written-code/ Sun, 27 Sep 2026 09:00:00 +0000 Coding agents write a great deal of code, quickly and fluently. Some of it is insecure. That is not because AI is uniquely careless. https://ives-yim.org.uk/writing/the-ai-supply-chain/ https://ives-yim.org.uk/writing/the-ai-supply-chain/ Sun, 27 Sep 2026 09:00:00 +0000 Engineering companies understand supply chains. We know that a product is only as reliable as its components, that a supplier's change can ripple through everything built on it, and that knowing exactly what is in a product is the foundation of quality, safety and compliance. https://ives-yim.org.uk/writing/the-attack-surface-has-moved/ https://ives-yim.org.uk/writing/the-attack-surface-has-moved/ Sun, 27 Sep 2026 09:00:00 +0000 For most of my career, securing a system meant understanding its boundaries. Where does data come in? Who can reach which interface? What runs with which privileges? https://ives-yim.org.uk/writing/the-hobby-system-and-the-professional-system/ https://ives-yim.org.uk/writing/the-hobby-system-and-the-professional-system/ Sun, 27 Sep 2026 09:00:00 +0000 Why "it works" is where the real engineering starts https://ives-yim.org.uk/writing/vibe-coding-is-another-abstraction/ https://ives-yim.org.uk/writing/vibe-coding-is-another-abstraction/ Sun, 27 Sep 2026 09:00:00 +0000 When I started as an engineer, I wrote real-time software in assembler for avionics. Each instruction mapped almost directly to what the processor did. Then higher-level languages took over, then operating systems and libraries, then frameworks, then cloud platforms. https://ives-yim.org.uk/writing/what-an-agent-actually-is/ https://ives-yim.org.uk/writing/what-an-agent-actually-is/ Sun, 27 Sep 2026 09:00:00 +0000 "Agent" has become the most overused word in technology. I have seen it applied to a chatbot with a company logo, a scheduled script that calls a language model, and systems that genuinely plan, act and adapt on their own. When one word covers all of those, it stops helping anyone make decisions. https://ives-yim.org.uk/writing/what-tokens-are/ https://ives-yim.org.uk/writing/what-tokens-are/ Sun, 27 Sep 2026 09:00:00 +0000 Anyone who uses coding agents seriously soon meets the word "token". It appears on the bill, in the limits, and in every discussion of why an agent slowed down or lost track of what it was doing. https://ives-yim.org.uk/writing/when-the-data-gives-the-orders/ https://ives-yim.org.uk/writing/when-the-data-gives-the-orders/ Sun, 27 Sep 2026 09:00:00 +0000 Imagine a support assistant that reads incoming customer tickets and helps your engineers diagnose faults. It has read-only access to telemetry and build records, as described in From what the manual says to what the product is doing . https://ives-yim.org.uk/writing/where-should-your-ai-live/ https://ives-yim.org.uk/writing/where-should-your-ai-live/ Sun, 27 Sep 2026 09:00:00 +0000 Over the past year, AI has become the most productive tool I have ever used as an engineer. I have used it to design, build, test and review systems that would previously have needed a team. https://ives-yim.org.uk/writing/workflows-before-agents/ https://ives-yim.org.uk/writing/workflows-before-agents/ Sun, 27 Sep 2026 09:00:00 +0000 When organisations ask me about AI agents, the most useful thing I can usually tell them is that what they need is not an agent. It is a workflow with AI in it. https://ives-yim.org.uk/writing/your-company-already-knows-the-answer/ https://ives-yim.org.uk/writing/your-company-already-knows-the-answer/ Sun, 27 Sep 2026 09:00:00 +0000 In every engineering organisation I have worked in, the most reliable way to find something out was to ask the person who had been there longest. The information existed somewhere: in a manual, a service report, a test log, an email thread from four years ago. https://ives-yim.org.uk/writing/when-the-product-is-already-in-the-field/ https://ives-yim.org.uk/writing/when-the-product-is-already-in-the-field/ Mon, 18 May 2026 09:00:00 +0000 Once a connected product ships, every false economy comes back through real units, customers and service. https://ives-yim.org.uk/writing/what-a-cto-should-be-hearing/ https://ives-yim.org.uk/writing/what-a-cto-should-be-hearing/ Fri, 15 May 2026 09:00:00 +0000 The signal a CTO most needs sits outside engineering, in four functions most never listen to. https://ives-yim.org.uk/writing/when-ai-removes-the-translation-layer/ https://ives-yim.org.uk/writing/when-ai-removes-the-translation-layer/ Fri, 15 May 2026 09:00:00 +0000 AI did not build the platform. It removed the handoff between my decisions and working code. https://ives-yim.org.uk/writing/not-all-ctos-are-created-equal/ https://ives-yim.org.uk/writing/not-all-ctos-are-created-equal/ Tue, 04 Nov 2025 09:00:00 +0000 The CTO title hides three different jobs. Early-stage companies need to know which one they are hiring for. https://ives-yim.org.uk/writing/connected-data-disconnected-decisions/ https://ives-yim.org.uk/writing/connected-data-disconnected-decisions/ Fri, 16 May 2025 09:00:00 +0000 Connected product platforms are repeating the failures of enterprise knowledge management. https://ives-yim.org.uk/writing/the-ota-problem/ https://ives-yim.org.uk/writing/the-ota-problem/ Fri, 16 May 2025 09:00:00 +0000 Over-the-air update is not just a deployment mechanism. It is product strategy, compliance and risk. https://ives-yim.org.uk/writing/what-the-launch-actually-costs/ https://ives-yim.org.uk/writing/what-the-launch-actually-costs/ Fri, 16 May 2025 09:00:00 +0000 Shipping is not launching. A connected product is only launched when the field can support it. https://ives-yim.org.uk/writing/turning-confusion-clarity-tackling-everybodys-job-nobodys/ https://ives-yim.org.uk/writing/turning-confusion-clarity-tackling-everybodys-job-nobodys/ Thu, 30 May 2024 09:00:00 +0000 A struggling company is rarely short of effort. It is short of ownership, and the fix starts at the top. https://ives-yim.org.uk/writing/cost-of-skipping-product-strategy/ https://ives-yim.org.uk/writing/cost-of-skipping-product-strategy/ Thu, 11 Apr 2024 09:00:00 +0000 Building before the strategy is defined turns a product launch into an expensive retrofit. https://ives-yim.org.uk/writing/c-suite-a-startup-actually-needs/ https://ives-yim.org.uk/writing/c-suite-a-startup-actually-needs/ Wed, 10 Apr 2024 09:00:00 +0000 The question is not which roles a startup needs. It is which role it needs next, and when. https://ives-yim.org.uk/writing/how-to-diagnose-strategic-deficit/ https://ives-yim.org.uk/writing/how-to-diagnose-strategic-deficit/ Thu, 04 Apr 2024 09:00:00 +0000 In connected products, direction becomes hardware. Analyse first, and define failure before you start. https://ives-yim.org.uk/writing/when-a-company-has-no-strategy/ https://ives-yim.org.uk/writing/when-a-company-has-no-strategy/ Thu, 04 Apr 2024 09:00:00 +0000 Strategic deficit shows up as a cluster of symptoms, and engineering often feels it before anyone names it. https://ives-yim.org.uk/writing/metrics-that-hide-the-failure/ https://ives-yim.org.uk/writing/metrics-that-hide-the-failure/ Wed, 03 Apr 2024 09:00:00 +0000 The interesting failures sit between domains. Measure the boundaries, not just the parts. https://ives-yim.org.uk/writing/where-technical-debt-becomes-expensive/ https://ives-yim.org.uk/writing/where-technical-debt-becomes-expensive/ Wed, 03 Apr 2024 09:00:00 +0000 Technical debt costs most in hardware, in architecture and at the boundaries between teams. https://ives-yim.org.uk/writing/what-downsizing-does-to-a-company/ https://ives-yim.org.uk/writing/what-downsizing-does-to-a-company/ Wed, 13 Mar 2024 09:00:00 +0000 A downsizing removes knowledge as well as cost, and that loss only shows once it has gone. https://ives-yim.org.uk/writing/leading-without-pretending-to-know/ https://ives-yim.org.uk/writing/leading-without-pretending-to-know/ Mon, 11 Mar 2024 09:00:00 +0000 Honesty about uncertainty is not a weakness in a CTO. Performed confidence is. https://ives-yim.org.uk/writing/art-of-reliable-embedded-systems/ https://ives-yim.org.uk/writing/art-of-reliable-embedded-systems/ Fri, 08 Mar 2024 09:00:00 +0000 Embedded reliability is decided early, in hardware, architecture and process, before the field finds out. https://ives-yim.org.uk/writing/what-it-takes-to-be-an-outstanding-cto/ https://ives-yim.org.uk/writing/what-it-takes-to-be-an-outstanding-cto/ Wed, 28 Feb 2024 09:00:00 +0000 Depth that transfers, breadth that informs judgement, and staying close enough to the work to use both. https://ives-yim.org.uk/writing/what-vision-means-in-a-cto/ https://ives-yim.org.uk/writing/what-vision-means-in-a-cto/ Wed, 28 Feb 2024 09:00:00 +0000 Vision in a CTO is not prediction. It is reading the constraints already present in a system. https://ives-yim.org.uk/writing/ethics-problem-in-ai-decision-making/ https://ives-yim.org.uk/writing/ethics-problem-in-ai-decision-making/ Mon, 26 Feb 2024 09:00:00 +0000 When hardware in the field acts on an AI classification, human oversight has to be engineered, not assumed. https://ives-yim.org.uk/writing/unlocking-brilliance-how-embracing-neurodiversity-fuels/ https://ives-yim.org.uk/writing/unlocking-brilliance-how-embracing-neurodiversity-fuels/ Wed, 21 Feb 2024 09:00:00 +0000 Organisations that only hear neurotypical thinking waste capacity. Better structure recovers it for everyone.