<?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 Prashant Kelker on Medium]]></title>
        <description><![CDATA[Stories by Prashant Kelker on Medium]]></description>
        <link>https://medium.com/@prkelker?source=rss-17ff44ea2930------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*0-0ydUXJgstaHCBzVovrOA@2x.jpeg</url>
            <title>Stories by Prashant Kelker on Medium</title>
            <link>https://medium.com/@prkelker?source=rss-17ff44ea2930------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 19:42:11 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@prkelker/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[Stop implementing software systems, chase business outcomes instead]]></title>
            <link>https://medium.com/rewiring-for-digital/stop-implementing-software-systems-chase-business-outcomes-instead-8776e2955d5c?source=rss-17ff44ea2930------2</link>
            <guid isPermaLink="false">https://medium.com/p/8776e2955d5c</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[architecture]]></category>
            <category><![CDATA[digital-transformation]]></category>
            <category><![CDATA[transformation]]></category>
            <category><![CDATA[digital]]></category>
            <dc:creator><![CDATA[Prashant Kelker]]></dc:creator>
            <pubDate>Sat, 02 Jan 2021 18:32:48 GMT</pubDate>
            <atom:updated>2021-01-02T18:32:48.721Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XA7D3mea-p4zk_2zoLvOJw.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@hishahadat?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Shahadat Rahman</a> on <a href="https://unsplash.com/s/photos/systems?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>You could have heard a pin drop in the room. We were debating the target architecture of a large program.</p><p>Someone in the room had said: “We had a target architecture four years ago — which we all believed in. And I guess we are half-way through implementing it. Why is this not helping us achieve the outcomes we are chasing now? It’s because everything is morphing — the business, the situation, the customer, the outcomes….”</p><p>We are distracted by new technology and solutions.</p><h4>Stop trying to implement a system — it is distracting us from our original goal</h4><p>Most IT discussions at enterprise level focus on what kind of system or architecture is required to achieve all the business goals that they have set out for themselves. The dialog then quickly moves to what software one should use, or how fast the target architecture can be implemented, or even what the right mix between commercial and custom built software should be.</p><p>One never gets to discussing what business outcomes the company was trying to achieve in the first place. Why is this such a hard discussion to have?</p><p><em>Probably because it is easier to talk about how to design a system, than to chase the elusive business outcome.</em> Business Outcomes are a moving target — moving constantly because the needs and demographics of the customer is continuously morphing. Designing a system is an in-house discussion — a discussion that is easier to have — without stepping out of the office.</p><p>Therein lies the conundrum: <em>How does one develop software for a goal that is changing constantly?</em></p><p>Some successful endeavours in the financial and retail industry over the last 12 months show promise — but the patterns that emerge are quite contrarian.</p><h4>Invest in winning teams, not in promising projects</h4><p>Winning teams who chase constantly moving targets and achieve continuous business outcomes do not plan large programs. A partnership with a new startup, a deal struck between two companies, an idea for a campaign in Denmark — this is what continuously achieving success looks like. Except that designing software for this new way of working goes against the grain of everything that we have been taught till date on what good software engineering should look like.</p><h4>Move away from projects &amp; programs to a continuous-state-of-achieving</h4><p>Such winning teams constantly spawn ideas and go from one victory to another. A smart enterprise should fund such ideas and track each idea’s path to victory. This turns program and portfolio management on its head. Suddenly <em>it is much more about investing in ideas that work, rather than guess upfront what projects one should sponsor.</em> There are multiple patterns from the venture capital industry that one can learn from to implement this way of funding and releasing capital.</p><p>A large retailer in Europe has moved to this new way of working — and continuously implemented micro solutions in the market. Each “micro-solution” gets funded in stages after it demonstrates that it can move the needle on a business KPI. With this, the team is forced to go beyond just going live successfully with a system to actually demonstrating a success in the market.</p><h4>Enterprise Architecture is an exercise in constant learning</h4><p>Well meaning enterprise architects have been trying, over years, to capture the essence of an organisation into neat boxes called capabilities, and designed and mapped these capabilities to software. But this is ignores the fact that we are creating an architecture for a moving target, and using solutions that have not yet taken the latest technology platforms into consideration. Five years ago, one would know about containerisation only if one were a Google employee. In the light of such technology changes, can one really hope to capture all that one needs over the next five years in a target architecture?</p><p>Instead we should focus on how the constant stream of solutions that create business outcomes evolve within a constantly morphing architecture. And enterprise architecture is then about the art of building this architecture as it evolves, rather than guessing what one would need.</p><h4>Focus on delivering benefits, instead of a system or scope</h4><p>Good software is a process, not a deliverable. Let us not allow technology and process to distract us. Nor allow decades of software architecture experience to blind us to the more agile way of achieving business outcomes.</p><p>Let us instead learn how to find and enable winning teams with full-stack responsibility without needing to change the organisation structure. Let us learn to live with the redundancy in solutions that comes as an unavoidable cost while chasing three simultaneous business outcomes. Let us focus our architectural talents instead on learning to deal with an evolving solution architecture built on technology that has been shaped by business decisions.</p><p>And above all, let us stop looking for systems to implement.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8776e2955d5c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/rewiring-for-digital/stop-implementing-software-systems-chase-business-outcomes-instead-8776e2955d5c">Stop implementing software systems, chase business outcomes instead</a> was originally published in <a href="https://medium.com/rewiring-for-digital">Rewiring for Digital</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Five ways to scale business strategy with AI]]></title>
            <link>https://medium.com/rewiring-for-digital/five-ways-to-scale-business-strategy-with-ai-e371536c0c08?source=rss-17ff44ea2930------2</link>
            <guid isPermaLink="false">https://medium.com/p/e371536c0c08</guid>
            <category><![CDATA[strategy]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[partnerships]]></category>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Prashant Kelker]]></dc:creator>
            <pubDate>Sun, 17 Jun 2018 17:46:17 GMT</pubDate>
            <atom:updated>2018-06-17T17:46:17.700Z</atom:updated>
            <content:encoded><![CDATA[<p>The enterprise world woke up to Artificial Intelligence (AI) in 2016. An AI software called Benjamin wrote the screenplay of a sci-fi movie. This movie “Sunspring” screened in the Sci-Fi-Film festival in London in summer 2016. The supercomputer Alpha Go beat Lee Sedol, the reigning world champion in Go. The potential of AI had arrived.</p><p>Ever since, Fortune 500 companies have started to grapple with what AI can mean to their business and their operations.</p><p>And as they experiment, a clear pattern is emerging. Anything which is a process can and will be taken over by AI &amp; machine learning (ML). If one extrapolates this, then all normal processes and cases will be automated and the job of humans will be relegated to handling the complex exceptions.</p><p>A recent HBR [<a href="https://hbr.org/2018/05/stitch-fixs-ceo-on-selling-personal-style-to-the-mass-market">article</a>] spoke about how at Stitch Fix, an online clothing subscription service, thousands of human stylists work with an AI recommendation engine to make personalised suggestions for customers. The machines enable the stylists with the speed for productivity, while the stylists provide the additional judgment needed for accurate recommendations.</p><p>As with any emerging technology, the only thing that is standing in the way of leveraging AI / ML is neither the technology nor its implementation. Instead, there are five factors that will inhibit or scale the growth of an organisation using AI/ML: The ability to imagine, the willingness to change, the rethinking of product design, mastering the art of partnering, and finally viewing AI as a paradigm for business change.</p><p><strong>Critical Skill 1: Unleash the imagination of your workforce</strong><br>The crux in AI &amp; ML does not lie in the lack of technology, but rather the conjuring up of use cases. How do we enable each person in the organisation to think big and imagine what the future will be like, and picture him/herself in this picture. What is blocking this imagination? One of the culprits is the hierarchical organisation structures that inadvertently shape the mindsets of people and narrow their thinking. The creation and nurturing of communities within the organisation will enable talent and creativity to cross silos and organisational boundaries.</p><p><strong>Critical Skill 2: Overcome the willingness to change</strong><br>It is counterproductive to force intelligent technology down the throat of the entire enterprise. Instead, involve employees in designing the balance between investment in AI &amp; ML, and maintaining the existing business.</p><p>Contrary to expectations, employees are far more willing to master new technology as they know that this will keep them employed.</p><blockquote>Redesign business processes from end to end, instead of a piecemeal approach. This sets the human factor in context with AI/ML.</blockquote><p>This will unleash a new set of human talents that machines are currently unable to match — creativity, interpersonal communication and empathy. Leading to the creation of new jobs and responsibilities.</p><p><strong>Critical Skill 3: Rethink existing product design &amp; architecture patterns</strong><br>Technologists have been designing deterministic systems since decades. Software products and systems are designed based on clarity of inputs and outputs. As we approach the AI world, systems learn how to deal with data through self-learning algorithms. And the art no longer lies in how to build systems with clear rules and logic. New skills will be required to build systems that constantly learn as more data is thrown at them. This will fundamentally reboot software development paradigms like requirement analysis, design and testing — whether done agile or waterfall.</p><blockquote>The world of ERP systems will change — customising packaged software to model the processes of an organisation will no longer be required, as AI combined with the rapid creation of apps will relegate ERP systems into operating engines. Data architectures will come to the forefront.</blockquote><p><strong>Critical Skill 4: Master the art of scouting &amp; partnering</strong><br>Working with providers and technology platforms will undergo a significant change as we leave the realm of deterministic systems. Organisations will have to learn to rely on providers as part of the AI solutions are embedded in their offerings. The discussion will move away from “which partner delivers the best quality” . As AI is applied to each business process, companies will have to ask themselves whether AI in this case is a core competence or “whether a partner can be trusted to run AI ethically and competently within a business process”. Startups in the area of AI have sprung up like mushrooms — each having a narrow focus, lack of real investment capital and the access to the enterprise market. This is a huge matchmaking opportunity.</p><blockquote>Procurement departments have honed the art and skill of either buying or owning assets and are yet to sharpen their axes in the skill of partnering.</blockquote><p>Partnering involves contracting for co-creation, joint success and sharing of the joint intellectual property created. Fortune 100 enterprises have fallen behind in the art of partnering at enterprise grade.</p><p><strong>Critical Skill 5: View AI as a paradigm for business design</strong><br>As the use cases of AI explode, organisations cannot afford to wait till first-movers in the industry show what is possible. The days of aiming to be the fast follower are over. Yet the pressure to act will depend a lot on how competitive the industry segment is. Even so, one can act ahead of the curve as industry segments merge. Manufacturing companies are tracking finance and retail companies in using AI in marketing. Insurance companies are adopting manufacturing paradigms like Digital Twins to predict the risk of the assets they insure. AI therefore should be seen not only as the means to improve the operations of an organisations, but rather fundamentally change the way that an organisation makes its revenue.</p><blockquote>A European Insurance firm Talanx uses learning systems that scour the internet for news releases of car accidents to detect fraudulent claims. This sounds easy till one thinks of the impact this will have on rudimentary downstream topics like, for instance, the KPIs of its salesforce or the design of payment systems.</blockquote><p><em>In conclusion: </em>AI and ML have the power to redefine value chains globally. Market leaders should look at AI as a means of defending their pole positions, while others should use AI to reposition themselves during the disintegration of the value chain. The above five critical skills will be key in whether AI will play a role in the strategy execution of an organisation.</p><p>This will not be an easy ask — it will require the extrapolation from data and algorithms to concrete product ideas and a forced investment even when the returns and goal horizons are not yet clear.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e371536c0c08" width="1" height="1" alt=""><hr><p><a href="https://medium.com/rewiring-for-digital/five-ways-to-scale-business-strategy-with-ai-e371536c0c08">Five ways to scale business strategy with AI</a> was originally published in <a href="https://medium.com/rewiring-for-digital">Rewiring for Digital</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Can we go from Doing Agile to Being Agile?]]></title>
            <link>https://medium.com/rewiring-for-digital/can-we-go-from-doing-agile-to-being-agile-919f0e1b31c1?source=rss-17ff44ea2930------2</link>
            <guid isPermaLink="false">https://medium.com/p/919f0e1b31c1</guid>
            <category><![CDATA[organization-design]]></category>
            <category><![CDATA[organizational-agility]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[agility]]></category>
            <dc:creator><![CDATA[Prashant Kelker]]></dc:creator>
            <pubDate>Sat, 09 Dec 2017 13:38:52 GMT</pubDate>
            <atom:updated>2017-12-09T13:38:52.293Z</atom:updated>
            <content:encoded><![CDATA[<p>2017 will probably go down in history as the year where the word “agile” entered mainstream enterprise lingo. And this was long overdue and laudable.</p><p>What is sad, however, is to see many enterprises going down the road of embracing “agile” as a process.</p><p>And what makes things worse? A lot of commonly found literature addresses agility at the level of a team or at most a department with a common goal. And G2000 organisations are attempting to take scaled versions of these ideas to apply at enterprise level — embracing LeSS, SAFe,….</p><p>But addressing the dimension of process is not enough.</p><p>In the quest to be more agile, top management teams in the G2000 now spend hours staring at Jira boards in their meetings, and have replaced the word “meeting” with “retrospective”. This must be frustrating.</p><p>Will embracing Agile ceremonies without a structure really change the culture? Is this really what we wanted? When did we go astray from “Being Agile” to “Doing Agile”?</p><blockquote>Structure &amp; Culture go hand in hand — and can feel iterative. Process can only help.</blockquote><p>Hopefully in 2018, we will start asking ourselves: “how do we structure ourselves for enterprise agility?”</p><h3>Structuring for Enterprise Agility</h3><p>We are building our “enterprise way of working” on projects. This is the legacy that we have inherited from Plan-Build-Run thinking. Over time, projects turn into task forces which then turn into structure.</p><p>Projects are discretionary and transitionary by nature. If we are structured around projects, this is like building out of blocks of ice that disappear over time and evaporate. Projects were never meant to be a structural component. Rob England makes a <a href="http://www.itskeptic.org/content/project-wave-product-structure">soul-searching case</a> on this topic.</p><p>And this gets even worse. Organisations have perfected the art of delivering projects and organised themselves around it in the form of committees and gated and phased delivery processes. This has led to teams that have in turn perfected the skill of developing and testing code and working themselves into silos.</p><p>Agile was a way to break this mold of thinking — but it has been implemented in the wrong way in G2000 organizations.</p><p>The same organisation that was structured around projects and activities has now hijacked the word “agile” and translated this into processes for agile — paving the way for SCRUM, LeSS and Scaled Agile Frameworks. But this is missing the point.</p><p>The question that needs to get answered is not “how do we work” as that can come later — the question that first needs answering is “which operating model best captures and encourages and business/market aligned delivery?” And therefore what should the guiding dimension for a standing team be?</p><p>One of the largest banks in Europe has been on a transformation journey to structure its more than 20000 IT employees into a product portfolio driven organisation structure. This massive transformation took more than three years to complete — resulting in 80 product families, and over 200 products. An interesting aspect is that they did not attempt to design the entire product portfolio sitting in an ivory tower, but instead dispersed the autonomy over the entire workforce.</p><blockquote>The enterprise architecture team gave the guiding direction on the structural product portfolio definitions — but left the individual teams to answer four key questions:</blockquote><blockquote>How does the usage (central, global, local, global + local) of your products affect the way you operate?</blockquote><blockquote>What does the above usage mean to your product portfolio in terms of landscape and support?</blockquote><blockquote>What would the modular structure below each product look like?</blockquote><blockquote>What parts of your portfolio are growing? Which products are entering retirement and need a “keep the lights on” mode?</blockquote><p>Product portfolio teams were then given the authority to discover what model works best for their individual product portfolios. We observed over time that some product teams restructured themselves two to three times over a span of 6–9 months till they got the best modular structure fit to their needs.</p><p>Not everything went smoothly and the transformation journey had to be replanned and interrupted several times to make way for the implementation of critical implementations and regulatory features. This bank had reached the right balance of centralising structural autonomy (what should our product portfolio be) vs operational autonomy (how should each product portfolio team structure itself to meets its varying needs).</p><h3>Going from Doing Agile to Being Agile</h3><p>But we should not fall victim to the Conway’s Law Trap: “organizations which design systems … are constrained to produce designs which are copies of the communication <a href="https://en.wikipedia.org/wiki/Organizational_structure">structures</a> of these organizations.”. Eric Raymond, an open source proponent, famously wrote: “If you have four groups working on a compiler, you will get a 4-pass compiler”</p><blockquote>So in 2018, can we really go from Doing Agile to Being Agile? Because this is so much more than just “ SCRUM, Agile Ways of Working and Agile Development”.</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=919f0e1b31c1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/rewiring-for-digital/can-we-go-from-doing-agile-to-being-agile-919f0e1b31c1">Can we go from Doing Agile to Being Agile?</a> was originally published in <a href="https://medium.com/rewiring-for-digital">Rewiring for Digital</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>