<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Ian Pattison on Medium]]></title>
        <description><![CDATA[Stories by Ian Pattison on Medium]]></description>
        <link>https://medium.com/@ianpattison2?source=rss-99416fc478df------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*a_LK0V2iOIYlpFh_4bdOAg.jpeg</url>
            <title>Stories by Ian Pattison on Medium</title>
            <link>https://medium.com/@ianpattison2?source=rss-99416fc478df------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 01:33:59 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ianpattison2/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[The Knowledge That Walks Out With Your Best People]]></title>
            <link>https://medium.com/@ianpattison2/the-knowledge-that-walks-out-with-your-best-people-277ea3fda041?source=rss-99416fc478df------2</link>
            <guid isPermaLink="false">https://medium.com/p/277ea3fda041</guid>
            <category><![CDATA[small-business]]></category>
            <category><![CDATA[future-of-work]]></category>
            <category><![CDATA[knowledge-management]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[leadership]]></category>
            <dc:creator><![CDATA[Ian Pattison]]></dc:creator>
            <pubDate>Tue, 16 Jun 2026 06:01:00 GMT</pubDate>
            <atom:updated>2026-06-16T06:01:00.798Z</atom:updated>
            <content:encoded><![CDATA[<h4>Agentic AI can capture what your best people know before they leave. But only if you ask the right questions first</h4><figure><img alt="an open office door" src="https://cdn-images-1.medium.com/max/1024/1*8UuS5qgDux2vqS9zy4BcNA.png" /></figure><p>Every business has a person like this. The one who knows which supplier to call when the usual one fails. The one who can look at a margin report and tell you, without a spreadsheet, whether the number is wrong. The one whose exit, when it finally comes, leaves a gap that no job advert fully fills. For most of the history of business, that knowledge walked out with them. Agentic AI changes that equation — but only if you move before they go.</p><p>The distinction worth making here is between documented process and tacit knowledge, and it is a distinction that most businesses have never had a strong reason to care about until now. Documented process is the procedure manual, the training guide, the policy document. Tacit knowledge is everything else: the judgment calls that the procedure manual never anticipated, the exceptions that an experienced person handles without thinking, the contextual understanding of a customer relationship that exists only in one person’s head. When you automate a process, you can codify the documented part relatively straightforwardly. The tacit part is where most implementations quietly fail.</p><p>I have seen this pattern in consulting work across retail and travel. A business automates a customer query process and is satisfied with the result — until an edge case arrives that the agent cannot handle, because nobody ever wrote down how to handle it. The experienced person who would have known exactly what to do has retired, or moved on, or is on holiday. The agent escalates. The customer waits. The gap that looked invisible in testing is suddenly visible in production.</p><p>The answer is not to document more before you automate. The answer is to capture differently. Structured interview techniques — the kind used in knowledge engineering, adapted for a business context — are designed specifically to surface tacit knowledge from experts who cannot tell you what they know, because they have never had to. You do not ask someone to describe their job. You ask them to walk you through the last three times something went wrong, and what they did. You ask them what they would tell a very competent new colleague that they would not find in the handbook. You ask them to describe the moment they knew a situation was about to become a problem. The answers to those questions are the knowledge that matters.</p><p>There is a second layer that most businesses miss entirely, and it is the one that determines whether an agentic system is trustworthy or merely fast. Exception conditions: the moments where the right answer is not to process the case, but to stop and call a human. A good invoice-processing agent does not just reconcile invoices. It knows which invoices need a human decision before anything happens, and it knows this because someone, at the design stage, thought carefully about what those cases look like and built the tripwires in. Getting this right requires the same kind of structured conversation with your most experienced people. What are the cases you would never let a junior person handle alone? What are the signals that tell you a situation needs escalation? These questions are uncomfortable to ask, because they surface the weaknesses of your operation. They are also the most valuable questions you can ask before you automate anything.</p><p>The broader point is this. A business’s operational knowledge — the accumulated judgment of the people who know how it actually works, as opposed to how it is supposed to work — is an asset that almost never appears on a balance sheet and is almost never treated as one. Agentic AI makes that knowledge executable. It can run the process, apply the judgment, and escalate the exception, at scale and without the salary cost. But it can only do this if the knowledge has been captured. And the time to capture it is before the person who holds it leaves, not after.</p><p>The practical implication is straightforward: if you are planning to automate any business process with agentic AI, the knowledge capture conversation should happen before the technical build, not alongside it. Bring in your most experienced operators. Ask them the questions that surface the tacit knowledge and the exception map. Record, structure, and use that material as the foundation for the agent’s design.</p><p>Your best people are a competitive asset. With the right approach, the knowledge they carry can outlast them.</p><p>Which process in your business is most dependent on knowledge that lives in one person’s head?</p><p><em>I help SME leaders to replace mundane work with end-to-end agentic AI. If this piece is relevant to a decision you are facing, feel free to reach out on LinkedIn.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=277ea3fda041" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The 1990s Mistake SMEs Are Making Again]]></title>
            <link>https://medium.com/@ianpattison2/the-1990s-mistake-smes-are-making-again-229fbff264da?source=rss-99416fc478df------2</link>
            <guid isPermaLink="false">https://medium.com/p/229fbff264da</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[digital-transformation]]></category>
            <category><![CDATA[future-of-work]]></category>
            <category><![CDATA[small-business]]></category>
            <category><![CDATA[leadership]]></category>
            <dc:creator><![CDATA[Ian Pattison]]></dc:creator>
            <pubDate>Fri, 05 Jun 2026 11:12:03 GMT</pubDate>
            <atom:updated>2026-06-05T11:12:03.163Z</atom:updated>
            <content:encoded><![CDATA[<h4>I watched a British high-street giant dismiss the web in 1995. I am watching the same decision being made about agentic AI right now</h4><figure><img alt="an empty motorway at dusk" src="https://cdn-images-1.medium.com/max/1024/1*Cp-x1Km3Tl8ELFlPualprQ.png" /></figure><p>In 1995, I sat in a room with the senior leadership team of one of Britain’s best-known high-street retailers. The web was the topic. Someone had put together a short brief on what this new technology might mean for their business, and the response from around the table was, in essence: interesting, but not really for us. They were not stupid people. They were experienced, commercially successful, and completely wrong. I am hearing almost identical conversations today, thirty years later, about agentic AI.</p><p>The parallel is not a metaphor. It is the same decision, made by the same type of person, at the same point in an adoption curve, with the same justification: we are watching it carefully, we will move when the time is right, and we do not believe it is an existential threat to our business model. The companies that said this about the web in 1995 had, in most cases, a five-year window before the damage became irreversible. The window for agentic AI is shorter. Much shorter.</p><p>What makes agentic AI different from every other technology trend of the past decade is that it does not just change how work is done. It changes who does it, and at what cost. A business that deploys agentic AI across its back-office operations — invoice reconciliation, policy compliance, customer query handling, reporting — is not just more efficient than one that does not. It is structurally cheaper to run. And a structurally cheaper competitor is a different kind of threat from a faster or better-marketed one. You cannot out-feature your way past a cost base that is 30–40% lower than yours.</p><p>I spent years inside IBM watching the enterprise adoption of earlier AI systems, and one pattern repeated itself with uncomfortable regularity. The organisations that moved early did not do so because they had better information than the ones that moved late. They did so because they had a leader who had seen an adoption curve before and understood that the right moment to move is never when the technology is fully proven and the path is obvious. By that point, the advantage is already gone. The high-street retailer I mentioned in 1995 did eventually build a web presence. By the time they did, Amazon was a fact, not a threat on a slide.</p><p>The agentic AI window is open now, and it is open specifically for businesses in the £5M to £50M revenue band. This is a point worth dwelling on, because most AI commentary either targets the consumer or the enterprise, and SME leaders are left to translate case studies that do not fit their context. The translation, in this case, is straightforward. Agentic AI systems, built properly, do not require the infrastructure investment or the technical teams that enterprise deployments assumed. The practical threshold has moved. What required a team of engineers and six months of integration work two years ago can, in many cases, be operational in weeks. The barrier that justified waiting has largely gone.</p><p>What comes next, if the pattern holds, is a wave of AI-native startups built from the ground up without the legacy cost structures that established businesses carry. This is what happened with the web. The incumbents were not, for the most part, destroyed by other incumbents who moved faster. They were disrupted by companies that had never carried the old costs and did not have to unpick the old processes to adopt the new ones. The travel industry learned this. Retail learned it. The businesses in those sectors that survived and thrived were the ones that moved before the disruption was visible, not in response to it.</p><p>The question worth asking is not whether agentic AI will affect your business. It will. The question is whether you will be one of the companies that shapes what that looks like in your sector, or one of the companies that responds to what others have built.</p><p>I was in that room in 1995. I know how the story ended for the retailer around that table. I am not making predictions. I am reading a pattern I have seen before.</p><p>What decision are you watching being made in your business right now?</p><p><em>I help SME leaders to replace mundane work with end-to-end agentic AI. If this piece is relevant to a decision you are facing, feel free to reach out on LinkedIn.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=229fbff264da" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Rise of the AI-Stack Developer]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@ianpattison2/the-rise-of-the-ai-stack-developer-ff99c08c7878?source=rss-99416fc478df------2"><img src="https://cdn-images-1.medium.com/max/1792/1*0TE6y6CNv_bj6MFBT5ypnQ.png" width="1792"></a></p><p class="medium-feed-snippet">The AI stack developer is emerging as a critical new role in the modern tech industry</p><p class="medium-feed-link"><a href="https://medium.com/@ianpattison2/the-rise-of-the-ai-stack-developer-ff99c08c7878?source=rss-99416fc478df------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@ianpattison2/the-rise-of-the-ai-stack-developer-ff99c08c7878?source=rss-99416fc478df------2</link>
            <guid isPermaLink="false">https://medium.com/p/ff99c08c7878</guid>
            <category><![CDATA[full-stack]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[front-end-development]]></category>
            <category><![CDATA[software-development]]></category>
            <dc:creator><![CDATA[Ian Pattison]]></dc:creator>
            <pubDate>Tue, 01 Oct 2024 17:24:17 GMT</pubDate>
            <atom:updated>2024-10-01T21:47:12.014Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Using Google Earth Engine to Map Vegetation Health]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/climate-technology-magazine/using-google-earth-engine-to-map-vegetation-health-f2806a6a0b6?source=rss-99416fc478df------2"><img src="https://cdn-images-1.medium.com/max/2324/1*T3nF0vewIlz2ZhmPuwXPag.png" width="2324"></a></p><p class="medium-feed-snippet">In this post, we will look in more detail at how we can detect vegetation health using the satellite data available to us.</p><p class="medium-feed-link"><a href="https://medium.com/climate-technology-magazine/using-google-earth-engine-to-map-vegetation-health-f2806a6a0b6?source=rss-99416fc478df------2">Continue reading on Climate Technology Magazine »</a></p></div>]]></description>
            <link>https://medium.com/climate-technology-magazine/using-google-earth-engine-to-map-vegetation-health-f2806a6a0b6?source=rss-99416fc478df------2</link>
            <guid isPermaLink="false">https://medium.com/p/f2806a6a0b6</guid>
            <category><![CDATA[satellite-technology]]></category>
            <category><![CDATA[mapping]]></category>
            <category><![CDATA[climate-technology]]></category>
            <category><![CDATA[google-earth-engine]]></category>
            <category><![CDATA[geospatial]]></category>
            <dc:creator><![CDATA[Ian Pattison]]></dc:creator>
            <pubDate>Sun, 24 Sep 2023 13:28:42 GMT</pubDate>
            <atom:updated>2023-09-24T13:28:42.108Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Getting Started with Google Earth Engine]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/climate-technology-magazine/getting-started-with-google-earth-engine-9dcca31bd3f8?source=rss-99416fc478df------2"><img src="https://cdn-images-1.medium.com/max/1834/1*MUjA5ykUk9QR4v3YBGD7YA.png" width="1834"></a></p><p class="medium-feed-snippet">Ever wondered how they create those wonderful, multi-coloured satellite images showing natural processes at work in the world? Or do you&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/climate-technology-magazine/getting-started-with-google-earth-engine-9dcca31bd3f8?source=rss-99416fc478df------2">Continue reading on Climate Technology Magazine »</a></p></div>]]></description>
            <link>https://medium.com/climate-technology-magazine/getting-started-with-google-earth-engine-9dcca31bd3f8?source=rss-99416fc478df------2</link>
            <guid isPermaLink="false">https://medium.com/p/9dcca31bd3f8</guid>
            <category><![CDATA[mapping]]></category>
            <category><![CDATA[google-earth-engine]]></category>
            <category><![CDATA[climate]]></category>
            <category><![CDATA[geospatial]]></category>
            <category><![CDATA[satellite-technology]]></category>
            <dc:creator><![CDATA[Ian Pattison]]></dc:creator>
            <pubDate>Tue, 29 Aug 2023 13:07:48 GMT</pubDate>
            <atom:updated>2023-08-29T13:12:53.664Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Carbon Neutral: It’s not what you think it is]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/climate-technology-magazine/carbon-neutral-its-not-what-you-think-it-is-5f7d10a0ccfa?source=rss-99416fc478df------2"><img src="https://cdn-images-1.medium.com/max/1024/1*zsexJmtnNom29yL8iVatsw.png" width="1024"></a></p><p class="medium-feed-snippet">We&#x2019;ve all seen companies and services that claim to be &#x2018;carbon neutral&#x2019; or &#x2018;carbon negative&#x2019;, but what does it actually mean?</p><p class="medium-feed-link"><a href="https://medium.com/climate-technology-magazine/carbon-neutral-its-not-what-you-think-it-is-5f7d10a0ccfa?source=rss-99416fc478df------2">Continue reading on Climate Technology Magazine »</a></p></div>]]></description>
            <link>https://medium.com/climate-technology-magazine/carbon-neutral-its-not-what-you-think-it-is-5f7d10a0ccfa?source=rss-99416fc478df------2</link>
            <guid isPermaLink="false">https://medium.com/p/5f7d10a0ccfa</guid>
            <category><![CDATA[climate-technology]]></category>
            <category><![CDATA[sustainability]]></category>
            <category><![CDATA[greenhouse-gas-emissions]]></category>
            <category><![CDATA[climate-change]]></category>
            <category><![CDATA[carbon-emissions]]></category>
            <dc:creator><![CDATA[Ian Pattison]]></dc:creator>
            <pubDate>Wed, 16 Aug 2023 19:52:27 GMT</pubDate>
            <atom:updated>2023-08-16T19:52:27.160Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Remote Sensing: How It Works]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/climate-technology-magazine/remote-sensing-how-it-works-2a15688f76e?source=rss-99416fc478df------2"><img src="https://cdn-images-1.medium.com/max/1920/1*0BxbqtwGFXHA3k2FloiJnw.jpeg" width="1920"></a></p><p class="medium-feed-snippet">An introduction to remote sensing with satellite data, and how it can help to solve climate challenges</p><p class="medium-feed-link"><a href="https://medium.com/climate-technology-magazine/remote-sensing-how-it-works-2a15688f76e?source=rss-99416fc478df------2">Continue reading on Climate Technology Magazine »</a></p></div>]]></description>
            <link>https://medium.com/climate-technology-magazine/remote-sensing-how-it-works-2a15688f76e?source=rss-99416fc478df------2</link>
            <guid isPermaLink="false">https://medium.com/p/2a15688f76e</guid>
            <category><![CDATA[geospatial-analytics]]></category>
            <category><![CDATA[satellite-technology]]></category>
            <category><![CDATA[sustainability]]></category>
            <category><![CDATA[climate-change]]></category>
            <dc:creator><![CDATA[Ian Pattison]]></dc:creator>
            <pubDate>Wed, 02 Aug 2023 16:00:20 GMT</pubDate>
            <atom:updated>2024-09-26T15:03:33.564Z</atom:updated>
        </item>
    </channel>
</rss>