# Cannon Enterprises LLC - Site: https://cannonco.net/ ## How to Choose the First Workflow to Automate - Link: https://cannonco.net/how-to-choose-first-workflow-to-automate/ - Published: 2026-09-29T11:53:25-05:00 - Author: cannonco Most organizations do not have an automation problem. They have a workflow-selection problem. When teams start with the newest tool, the loudest complaint, or the most technically interesting process, they often automate complexity instead of removing it. A better first move is to choose one workflow that is repetitive, bounded, measurable, and important enough to matter without being so critical that experimentation creates unacceptable risk. ## Start with a workflow that repeats The best first automation candidates occur frequently enough that wasted effort is visible. Monthly reporting, intake triage, status follow-up, document routing, scheduling coordination, routine approvals, and recurring exception notifications are common examples. A workflow that happens once a year may still deserve automation, but it is usually a weaker first candidate because the organization gets fewer opportunities to learn, validate, and improve the automation. ## Choose a bounded process A useful first workflow has a recognizable beginning, a small number of systems or participants, and a clear end state. For example, “improve customer service” is too broad. “Route new service requests to the correct owner, notify the requester, and escalate requests that remain unassigned after two hours” is bounded enough to design, test, and govern. Boundaries matter because automation becomes fragile when every adjacent process is pulled into the first release. Keeping the initial scope narrow makes it easier to identify dependencies, permissions, data requirements, and human decision points. ## Look for measurable friction A strong automation candidate should improve something that can be observed. Useful baseline measures include cycle time, labor hours, backlog, error rate, rework, missed handoffs, response time, or the number of exceptions requiring executive attention. If the organization cannot describe what should improve, it will be difficult to determine whether the automation produced value. ## Keep accountable human decisions visible Automation should remove repetitive work, not accountability. Financial commitments, legal decisions, security exceptions, privacy-sensitive actions, unusual customer promises, and other material decisions should remain subject to defined human authority. The design question is not simply “Can this task be automated?” It is “Which steps should be automated, which steps require review, and what evidence should the process leave behind?” ## Prefer workflows with accessible data and systems A workflow is easier to automate when its information already exists in structured systems such as Microsoft 365, SharePoint, a CRM, a ticketing platform, WordPress, or another application with reliable connectors or APIs. If the process depends on undocumented tribal knowledge, inconsistent spreadsheets, or data that cannot be accessed securely, the first step may need to be process clarification or data cleanup rather than automation. ## Use a simple first-workflow test * Does the workflow happen frequently? * Is the start and finish easy to define? * Can current performance be measured? * Are the required systems and data accessible? * Can accountable human decisions remain clearly identified? * Would improving this workflow save time, reduce delay, or improve visibility? * Can the first release be tested without creating unacceptable operational risk? If most of those answers are yes, the workflow is a strong candidate for a first automation effort. ## Design before you build The fastest route to useful automation is often a short diagnostic before implementation. Cannon Automation Works’ [AI Workflow Diagnostic](https://cannonco.net/product/ai-workflow-diagnostic/) evaluates one defined business process, identifies the strongest automation opportunity, documents required human controls, and recommends a practical next step before a larger build is purchased. For organizations that already know which process they want to improve, the goal is not to automate everything. It is to choose the right first workflow, prove value, preserve accountability, and build from evidence. ## How to Estimate Automation ROI Before You Build - Link: https://cannonco.net/estimate-automation-roi-before-build/ - Published: 2026-09-28T09:55:01-05:00 - Author: cannonco Automation ROI should be understandable in operational terms before a build begins. A technical demonstration may be impressive, but value comes from measurable changes in the work. ## Start with the current workload Estimate how often the process runs, how many people touch it, and how much time each cycle consumes. Include coordination work such as status checks, reminders, re-entry of information, manual reporting, and exception follow-up. ## Measure friction, not only labor Manual effort is only one source of value. Delays, missed handoffs, inconsistent routing, incomplete information, rework, and slow escalation can create larger operational costs than the minutes spent performing a task. ## Identify what automation can actually remove Do not assume the entire process disappears. Estimate which steps can be automated, which steps can be accelerated, and which decisions must remain human. A credible business case uses the expected future-state workflow rather than applying an optimistic percentage to the whole process. ## Include implementation and operating cost Account for design, configuration, integration, testing, documentation, maintenance, monitoring, licensing, and exception handling. Simple automations can produce excellent returns precisely because their ongoing operating burden is low. ## Use a short list of operational measures Useful measures include cycle time, manual touches, overdue actions, response time, rework, exception rates, reporting effort, and time-to-decision. Pick measures that the process owner can observe before and after implementation. ## Make the go/no-go decision explicit A diagnostic should be allowed to conclude that a process should be redesigned, deferred, or left alone. Avoiding a low-value build is itself a return. Cannon Automation Works uses this logic in the [AI Workflow Diagnostic](https://cannonco.net/automation/ai-workflow-diagnostic/), a focused review of one process before a larger implementation commitment. See the broader [automation service model](https://cannonco.net/automation/) for the path from diagnostic through implementation and managed operations. ## What Should Stay Human in an AI Workflow? - Link: https://cannonco.net/what-should-stay-human-in-ai-workflow/ - Published: 2026-09-28T05:52:10-05:00 - Author: cannonco Useful automation is not defined by how much human work it removes. It is defined by whether routine work is automated while accountable judgment remains where it belongs. ## Separate processing from authority Many workflows contain both mechanical steps and consequential decisions. Collecting information, checking required fields, routing a request, assembling a briefing, sending a reminder, and recording status are often good automation candidates. Approving a legal commitment, resolving conflicting evidence, authorizing unusual spending, handling a security exception, or making a decision with material personnel impact usually requires accountable human review. ## Define escalation conditions before launch An autonomous workflow should know when to stop its normal path. Common escalation conditions include missing or conflicting data, exceptions above a defined threshold, privacy or security concerns, unusual customer impact, contractual ambiguity, and actions that would materially change cost, scope, or authority. ## Design for evidence Human review works best when the system presents the decision maker with the relevant context: what happened, what changed, what the automation already checked, which rule was triggered, and what decision is required. This reduces the burden on the reviewer without hiding responsibility. ## Keep the exit ramp visible Automation should have a clear way to pause, override, or revert. The operating model should identify who owns that authority and what evidence is retained when an exception occurs. ## Measure the right outcome The objective is not maximum autonomy. The objective is reliable operations. Measure cycle time, manual touches, exception volume, response time, rework, and the quality of the information presented to human decision makers. If you are deciding where automation can safely reduce work in one business process, review the [AI Workflow Diagnostic](https://cannonco.net/automation/ai-workflow-diagnostic/) and the [Cannon Automation Works human-oversight approach](https://cannonco.net/automation/ai-transparency/). ## 7 Signs a Business Process Is Ready for Workflow Automation - Link: https://cannonco.net/business-process-ready-for-workflow-automation/ - Published: 2026-09-28T02:53:11-05:00 - Author: cannonco Not every repetitive task is a good automation candidate. The best early targets are processes where manual coordination is consuming time, the operating rules are reasonably clear, and the organization can define what should happen when the normal path breaks. ## 1. The same work happens repeatedly Repetition creates leverage. Intake forms, status updates, document requests, recurring reports, lead routing, work-order coordination, approval reminders, and scheduled follow-up are common examples. ## 2. The handoffs are predictable A process becomes easier to automate when you can identify who receives the work, what information they need, and what event moves the work to the next step. ## 3. People spend time chasing status If employees repeatedly ask whether a request was received, who owns the next action, whether a document arrived, or when something is due, the process may have a coordination problem rather than a staffing problem. ## 4. The inputs can be defined Automation needs a reliable starting point such as a form, email, SharePoint item, CRM record, approved document, scheduled event, or API signal. ## 5. Human decision points can be separated from routine work A strong design distinguishes routine processing from decisions that require judgment, authority, or review. High-impact exceptions should have an explicit human escalation path. ## 6. Success can be measured operationally Useful measures include cycle time, manual touches, overdue actions, exception rates, response time, rework, reporting effort, or time spent assembling information. ## 7. The process has an owner Every production automation needs someone accountable for the underlying business process. ## Start with one bounded workflow Choose one workflow with clear pain, visible repetition, defined participants, and measurable outcomes. Map the current process, identify the human control points, determine what systems are involved, and decide whether to automate, redesign, defer, or leave the process alone. Learn more about the [AI Workflow Diagnostic](https://cannonco.net/automation/ai-workflow-diagnostic/), the [Cannon Automation Works service model](https://cannonco.net/automation/), and our [approach to human oversight](https://cannonco.net/automation/ai-transparency/). ## Book Review: Mastering Knowledge Management Using Microsoft Technologies by Tori Reddy Dodla - Link: https://cannonco.net/book-review-mastering-knowledge-management-using-microsoft-technologies/ - Published: 2026-05-08T13:46:57-05:00 - Author: cannonco I recently went back through my blogs and emails and realized that I did not uphold a promise that I made to Tori back a couple of years ago. That is to review her book _[Mastering Knowledge Management Using Microsoft Technologies: Secrets to Leveraging Microsoft 365 and Becoming a Knowledge Management Guru](https://www.amazon.com/Mastering-Knowledge-Management-Microsoft-Technologies/dp/B0CZHP23LF/)_ by Tori Reddy Dodla. As someone who has spent a great deal of time working at the intersection of Knowledge Management, technology, governance, and organizational performance, I found this book both timely and useful. What I appreciated most is that the book addresses a very real challenge many organizations face today. Most organizations already have Microsoft 365, SharePoint, Teams, Power Platform, Power BI, and now Copilot somewhere in their environment. Yet many of those same organizations are still struggling with knowledge silos, inconsistent document management, poor search experiences, weak reuse of lessons learned, and fragmented collaboration practices. That is where this book provides value. Rather than treating Knowledge Management as an abstract concept or presenting Microsoft 365 as just another technology stack, Dodla brings the two together in a practical way. The book helps readers think through how the Microsoft ecosystem can support knowledge capture, organization, sharing, analysis, and reuse. For many KM practitioners, SharePoint administrators, digital transformation leads, and business leaders, that alone makes the book worth reading. The strength of the book is its practicality. It does not try to overcomplicate the conversation. It walks readers through how Microsoft tools can be used to build knowledge bases, manage documents, automate processes, visualize knowledge-related data, and improve access to organizational knowledge. The inclusion of Power Apps, Power BI, Power Platform, and Copilot makes the book especially relevant given where the workplace is heading. That said, my honest view is that this book is strongest as a technology-enabled KM implementation guide. It is not, and should not be viewed as, a complete enterprise Knowledge Management strategy. That distinction matters. Knowledge Management is not created simply because an organization has SharePoint. It is not solved by creating a Teams channel. It is not achieved by deploying Copilot. These tools can absolutely support KM, but they do not replace the need for strategy, governance, ownership, culture, taxonomy, metadata discipline, lessons learned processes, communities of practice, or leadership accountability. In other words, Microsoft 365 can enable Knowledge Management, but it cannot do the hard organizational work by itself. That is not a criticism of the book as much as it is a caution for the reader. If you are looking for a practical guide to better use the Microsoft tools your organization likely already owns, this book is a strong resource. If you are looking for a full KM maturity model, enterprise governance framework, or deep treatment of KM culture and behavior change, you will want to pair this book with broader KM literature and practical consulting guidance. Another point worth noting is that Microsoft technologies are changing quickly. Copilot, SharePoint Premium, Viva, Purview, Teams, and the Power Platform continue to evolve. Because of that, readers should treat the book as a strong foundation rather than a final technical playbook. The concepts will remain useful, but the specific technical steps should always be checked against the latest Microsoft guidance. Overall, I found _Mastering Knowledge Management Using Microsoft Technologies_ to be a timely, practical, and valuable contribution. It speaks directly to organizations that are trying to get more value from Microsoft 365 while improving how knowledge flows across the business. For KM professionals working in Microsoft-heavy environments, this book offers a helpful bridge between KM intent and technology execution. My rating would be **4 out of 5 stars**. It earns that rating because it is practical, relevant, and useful. I would not give it a full 5 because Knowledge Management is broader than any technology platform, and readers should be careful not to confuse Microsoft configuration with enterprise KM maturity. Still, for the right audience, this is a book I would recommend. ## Final Thought If your organization already uses Microsoft 365 but still struggles to find, trust, share, and reuse what it knows, this book is worth your time. Just remember that the technology is only part of the solution. The real work of Knowledge Management still comes down to people, process, governance, culture, and leadership. ## What Is Knowledge Management? - Link: https://cannonco.net/what-is-knowledge-management/ - Published: 2026-02-09T13:02:11-06:00 - Author: cannonco After more than two decades immersed in Knowledge Management across military, government, consulting, and corporate environments, I have learned one consistent truth: Knowledge Management is not a buzzword, and it is not a software deployment. Knowledge Management is a strategic discipline that helps organizations **learn faster than their environment changes**. It strengthens decision making, reduces preventable rework, improves mission and operational performance, and protects hard won expertise from walking out the door. If you want a simple definition that works in the real world: **Knowledge Management is the deliberate way an organization creates, curates, shares, applies, and improves what it knows, so people can make better decisions and deliver better outcomes.** That is the difference between “we store information” and “we operate with organizational intelligence.” --- ## The 5 Pillars of Knowledge Management Over time, I have seen countless frameworks and maturity models. Most work when they are grounded in execution. In practice, KM stands on five pillars that must reinforce each other. ### 1. People KM starts and ends with people. Tools do not share knowledge. People do. The highest performing environments make it easy and safe for people to: * Ask questions without penalty * Share lessons without blame * Teach others without losing status * Challenge assumptions with respect * Build networks of trust across teams When KM succeeds, it is because people are supported, recognized, and equipped to do knowledge work as part of normal work. **Practitioner signal:** If knowledge sharing depends on heroic volunteers, it will not scale. Make it part of the job, the rhythm, and the incentives. --- ### 2. Process Sustainable KM is built on repeatable processes that fit the operational cadence. KM processes are not extra work. They are how you reduce waste and improve performance. Examples that consistently produce results: * After Action Reviews and Retrospectives that lead to updates in guidance, training, and standards * Peer Assists before major work to reuse what already works * Communities of Practice that solve problems and standardize good practice * Structured onboarding and proficiency pathways that reduce time to competency * Knowledge capture for critical roles, not as “interviews,” but as operational handovers **Practitioner signal:** If you capture lessons but do not change how work is done, you have reporting, not learning. --- ### 3. Organizational Culture Culture is the true engine of KM. Where culture encourages openness and learning, KM thrives. Where culture rewards hoarding, KM becomes a checkbox. The cultural ingredients I see in strong KM organizations include: * Trust and psychological safety * Leaders who ask, “What did we learn?” and “What did we reuse?” * A norm that knowledge is an organizational asset, not personal property * A bias toward evidence, transparency, and continuous improvement Culture is not posters and slogans. Culture is what happens when deadlines hit and things go wrong. **Practitioner signal:** Watch what gets rewarded. If speed is rewarded but learning is punished, KM will always be fragile. --- ### 4. Tools and Technology **_“Technology is an enabler, not the solution”._** Platforms like SharePoint, ServiceNow, Confluence, and enterprise search can be powerful. Yet without the other pillars, they become expensive filing cabinets. In strong KM programs, tools are designed to make knowledge: * **Findable**: you can locate the best answer quickly * **Usable**: the content is written for action, not archives * **Trusted**: authoritative sources are clear, current, and governed * **Embedded**: knowledge appears in the workflow where decisions are made If users must hunt across multiple systems, KM adoption will collapse under real work pressure. **Practitioner signal:** Users do not want “more information.” They want the best answer, with confidence, at the moment of need. --- ### 5. Governance Governance is the backbone that makes KM durable. Governance is not bureaucracy. It is clarity. Good governance answers: * Who owns this knowledge domain * What is authoritative versus optional * How quality is reviewed and kept current * How access, privacy, and security are handled * How standards, taxonomy, and metadata are applied * How KM aligns to mission, strategy, and measurable outcomes Without governance, KM becomes a collection of well meaning efforts that drift, duplicate, and decay. **Practitioner signal:** If no one is accountable for knowledge quality and currency, your AI, analytics, and decisions will inherit that risk. ![What is KM Infographic](https://cannonco.net/wp-content/uploads/2026/02/What_is_KM_Graphic- 683x1024.png) --- ## KM Versus Information, Data, and Change Management A lot of organizations struggle because they blur these disciplines. Each is essential, but they are not the same. * **Data Management**: Manages raw data, definitions, lineage, quality, and stewardship. * **Information Management**: Organizes documents, records, content, and retrieval. * **Change Management**: Prepares people to adopt new ways of working and sustain behavior change. * **Knowledge Management**: Integrates people, process, culture, technology, and governance so information and experience become actionable insight and improved performance. A practical way to think about it: **Data becomes information when it is organized and contextualized. Information becomes knowledge when it is interpreted, shared, and applied.Knowledge becomes advantage when it drives better decisions and outcomes. > Example: SharePoint and ServiceNow can support information management very well. > The KM value appears when you add learning processes, communities, validated knowledge > assets, and governance that keep content accurate, discoverable, and operationally > relevant. --- ## The KM Outcomes Leaders Actually Care About KM is not measured by the number of documents uploaded or pages viewed. Those are activity metrics, not outcome metrics. KM creates value when it improves results such as: * Faster onboarding and reduced time to proficiency * Fewer repeat incidents and fewer preventable failures * Increased reuse of proven practices and reduced rework * Better decision quality through traceable rationale and evidence * Greater resilience during turnover, reorgs, or surge operations * Increased innovation by connecting expertise across silos If you cannot connect KM to operational outcomes, the program will always be vulnerable at budget time. --- ## KM in the Age of AI AI has raised the stakes. If your knowledge is ungoverned, outdated, duplicative, or difficult to trace back to authoritative sources, AI will amplify that problem. Many organizations are learning that AI readiness is less about the model and more about the knowledge foundation. In practice, modern KM must support: * Source grounded answers and citations to authoritative knowledge * Clear ownership and lifecycle management * Content quality controls and review rhythms * Taxonomy and metadata that improve retrieval precision * Decision records and rationale that support auditability AI does not replace KM. AI makes KM non negotiable. --- ## Closing Thought KM is not a one time initiative or a plug and play solution. It is a capability that grows with practice, leadership commitment, and disciplined execution. If your organization is ready to move beyond storing information and toward building real organizational intelligence, I would welcome the conversation. **Question for you:** Which pillar is the strongest in your organization today, and which one is the biggest constraint? --- ## The Allure of Technology - Link: https://cannonco.net/the-allure-of-technology/ - Published: 2025-11-11T17:47:45-06:00 - Author: cannonco “Don’t Buy the Tool Before You Know the Terrain: Why Every Organization Needs a Knowledge Assessment First” In today’s digital rush, organizations often jump head-first into implementing new technologies. AI systems, data lakes, collaboration platforms, or analytics dashboards, with CEO, CTO, and other C-Sutie officials believing that the right tool will automatically solve their knowledge problems. However, this approach can be likened to buying expensive gym equipment without ever assessing your fitness goals, capabilities, or habits. A recent client’s feedback captured this mindset succinctly: _“I was disappointed by the absence of technological integration in the proposal…”_ This sentiment reveals a common misconception that technology is strategy. Technology is only an enabler, not the foundation of effective Knowledge Management (KM). What a Knowledge Assessment Actually Does A Knowledge Assessment is a structured evaluation of how an organization creates, shares, stores, and applies knowledge to achieve its objectives. It identifies critical enablers and barriers across four domains: People, Processes, Technology, and Culture. Without this diagnostic phase, even the most advanced technology investments risk failing because they do not align with how knowledge flows within the organization. ISO 30401 (the international KM standard) emphasizes this alignment as the first step in building a sustainable KM system (ISO, 2018). Why Organizations Skip It Organizations often skip the assessment stage for three reasons: 1. Perceived urgency: leadership wants quick wins. 2. Vendor influence: solution providers market platforms as “plug-and-play.” 3. Budget optics: assessments are seen as overhead, not as value creation. The irony is that skipping the assessment usually costs far more in the long run. Research by Davenport and Prusak (1998) shows that up to 70% of knowledge initiatives fail when technology precedes strategy or assessment. Similarly, McKinsey (2020) found that firms that perform up-front knowledge audits see three times higher adoption rates for digital platforms. Lessons from the Field At Knoco, we have observed that organizations that neglect knowledge assessments often face predictable outcomes: * Misaligned technology that doesn’t fit business workflows. * Poor adoption because staff don’t see value. * Data duplication, versioning issues, and knowledge silos. * Reimplementation costs when the platform fails to deliver. The client quoted earlier, for example, prioritized technology over knowledge diagnostics. Six months later, their project stalled due to low engagement and unclear ownership. Symptoms that a proper Knowledge Assessment would have revealed early. The Hidden ROI of a Knowledge Assessment A Knowledge Assessment is not a cost; it’s a risk-mitigation investment. Its outcomes provide: * A baseline for measuring maturity and readiness. * A blueprint for aligning KM strategy to business goals. * A gap analysis identifying where people, processes, and technology must evolve. * A change management roadmap that ensures adoption, not just installation. Organizations that conduct these assessments typically reduce technology rework costs by 30–50% (Gartner, 2022) and experience higher user engagement within the first 90 days post-implementation. Technology Should Follow Knowledge The Knowledge Assessment informs us what technology is truly needed, not what looks modern or impressive. For instance: * If tacit knowledge is the biggest gap, invest in collaboration tools and communities of practice. * If explicit knowledge is poorly managed, strengthen content governance before upgrading platforms. * If decision latency is the issue, integrate AI-enabled analytics _after_ data and process alignment. The right tool will emerge naturally once the knowledge ecosystem is understood. Conclusion: Start with Knowledge, End with Intelligence Every digital transformation begins with a deceptively simple question: “Do we know what we know?” Without that clarity, no amount of technology can make an organization smarter. As the saying goes in Knowledge Management: _“You can’t automate what you don’t understand.”_ Conducting a Knowledge Assessment first ensures that technology becomes a multiplier of human intelligence, not a substitute for it. ## Global Knowledge Management Week 2025: A Snapshot - Link: https://cannonco.net/global-knowledge-management-week-2025-a-snapshot/ - Published: 2025-10-19T20:23:31-05:00 - Author: cannonco Every year globally connected KM practitioners, networks and organizations pause for a moment of alignment, sharing and celebration in what we now recognize as Global Knowledge Management Week (KM Week). This event provides a focal point for raising the visibility of knowledge management (KM) as a discipline, reinforcing its relevance across sectors and geographies, and encouraging coordinated activity, dialogue and reflection. #### Origins and Evolution The seed of this global “week of KM” belongs to the Knowledge Management Global Network (KMGN), a not-for-profit network of international KM communities. KMGN was founded in 2014 and institutionalized in 2021. ([kmglobalnetwork.org](https://www.kmglobalnetwork.org/)) Via KMGN and its partner communities, the notion of a “KM Week” or “KM Festival” emerged as a way to bring regional, national and organizational KM activities into a common timeframe and amplify their impact. For instance, KMGN describes “Global Knowledge Week” as being organized since 2023 under that banner. ([KMedu Hub](https://kmeducationhub.de/tag/kmgn-global-knowledge-week/)) More broadly, RealKM Magazine reports that in 2024, there were 138 events across 16 countries as part of the 2024 edition of KM Week — signaling meaningful traction and global spread. ([RealKM](https://realkm.com/2025/10/01/in-the-know-global-knowledge-week-2025-fundraising-for-institute-of-domestic-violence-religion-migration-idvrm/)) Thus, what began as a network-driven coordination effort has matured into an annual“ festival of KM” that invites knowledge professionals everywhere to plan events, share case studies, explore emerging topics (such as AI, tacit knowledge, knowledge graphs, decolonizing knowledge) and connect across borders. #### Why It Matters From my vantage, having worked in knowledge management in the military, government and private sectors, KM Week is powerful for several reasons: * It raises awareness of KM’s strategic importance in organisations and ecosystems( not just as a “nice-to-have” but as a value driver). * It creates a focal point for KM professionals to synchronise efforts, share best practices, and challenge themselves with new themes. * It elevates innovation and future-oriented thinking in KM (for example: how do generative AI, knowledge graphs, and contextual ontologies alter KM strategy?). * And it fosters a sense of global community among KM practitioners, which supports sharing across cultural, organisational and national boundaries. For someone working at the intersection of knowledge management, innovation and data management (in my case across defense, government and industry), KM Week serves as a useful anchor to reflect on where we stand as a KM Community, what is working, where the gaps are, and how to advance from “data → knowledge → decision” loops in increasingly complex environments. ### What is the Agenda for KM Week 2025? Here are the key details for this year’s edition, plus how organizations and practitioners might engage proactively. **Dates & Participation** * The 2025 edition of Global Knowledge Management Week is scheduled for **October 20-25, 2025**. ([ROM Global](https://www.kmrom.com/news-items/global-knowledge-management-week-2025/)) * The week is open: organizations, networks, communities and individuals are encouraged to plan events (online, in-person or hybrid) across regions, sectors and levels.( [ROM ](https://www.kmrom.com/news-items/global-knowledge-management-week-2025/) [Global](https://www.kmrom.com/news-items/global-knowledge-management-week-2025)) **Focus Themes** While the week itself is a “container”, a number of thematic tracks are emerging that reflect the current frontiers of KM. According to RealKM and other commentary: ([RealKM](https://realkm.com/2025/10/01/in-the-know-global-knowledge-week-2025-fundraising-for-institute-of-domestic-violence-religion-migration-idvrm/)) * **Generative AI & KM**: how AI agents, knowledge graphs and other “next-gen” tools intersect with KM practice. * **Tacit Knowledge & Knowledge Capture**: recognising that in many organisations critical knowledge remains tacit, embedded in human experience, often high-stakes as in defence or complex engineering. * **Complexity, Systems Thinking & KM**: acknowledging that knowledge flows in complex, dynamic systems and KM must adapt accordingly (e.g., project-based organizations, temporary organizations). * **Decolonizing Knowledge & Knowledge Sovereignty**: foregrounding culturally aware, inclusive approaches to KM that recognize diverse epistemologies and avoid extractive knowledge models. * **Knowledge Governance, Standards & Measurement**: as organisations mature in KM, questions of governance, ROI, standardisation (e.g., ISO 30401), and metrics become ever more relevant. * **Communities of Practice, Knowledge Sharing & Innovation**: emphasising collaboration, networks, peer-to-peer sharing and the link from KM to innovation outcomes. * **KM in SMEs & Varied Sectors**: KM does not just happen in large enterprises, need to look how KM practices adapt to smaller organizations, different industries( e.g., SMEs, project-based sectors, construction). **How to Get Involved – A Practitioner Checklist** For KM practitioners, leaders and innovation managers, here are suggested steps: 1. **Schedule a kickoff**: coordinate within your organisation (or network) a KM Week event, even a short webinar or panel, for the Oct 20–25 window. 2. **Select a theme**: pick one (or more) of the emerging tracks above that reflect your organisational context (e.g., AI-enabled knowledge capture in defence; tacit knowledge transfer in government). 3. **Invite cross-discipline input**: involve stakeholders beyond KM (e.g., data management, AI/ML, innovation teams, decision-makers) to broaden the conversation. 4. **Capture insights**: use the week as a “pulse check” to document what is happening, what gaps remain, what opportunities for improvement exist. 5. **Share out**: report key take-aways internally and externally (e.g., LinkedIn post, blog) to build momentum and visibility for KM in your ecosystem. 6. **Link to strategy**: tie the event back to your broader KM/innovation/data management strategy, this is not just a standalone event but a gateway to future action. 7. **Engage the global KM community**: connect with KMGN, share your event, and see what others are doing globally – the value lies in cross-pollination too. ### A Personal Note In over 20 years of working at the convergence of knowledge management, data management and innovation (especially in defense, government and industry), I have seen how historically knowledge functions were often relegated to support roles. But the accelerating pace of change, driven by digitalization, AI, networked operations and hybrid work, this means KM is now mission critical for all organizations. KM Week provides an ideal moment to step back, reflect on progress and lean into the future. As we enter KM Week 2025, I encourage you to treat it not just as a calendar event, but as a strategic opportunity: to align your knowledge agenda with innovation, put tacit knowledge into play, govern effectively, and connect to broader ecosystem- thinking. Your organization (and your network) will be better for it. Let us make this KM Week 2025 one of purposeful alignment, global connection and future-oriented action. ## Why Your AI is Only as Smart as Your Knowledge Management: The Hidden Foundation of Effective AI Implementation - Link: https://cannonco.net/why-your-ai-is-only-as-smart-as-your-knowledge-management-the-hidden-foundation-of-effective-ai-implementation/ - Published: 2025-08-06T20:32:56-05:00 - Author: cannonco # Introduction: The Myth of Plug-and-Play AI Artificial Intelligence (AI) is often sold as a turnkey solution—an omnipotent force capable of transforming organizations with minimal effort. Vendors promise streamlined operations, powerful predictions, and automated decision-making. Yet behind the scenes, many AI implementations fail to meet expectations. According to a 2023 Gartner report, **over 85% of AI projects never make it into production**, and those that do often fail to scale or deliver ROI. Why? Because **AI doesn’t work without knowledge.** And most organizations are sitting on fractured, inaccessible, or outdated knowledge assets. The uncomfortable truth is this: **your AI is only as smart as your Knowledge Management( KM) strategy and system.** In this blog, we’ll explore real-world case studies to expose how poor KM undermines AI—and how building a strategic KM foundation can supercharge your implementation. ## The AI Hype vs. the KM Reality While AI depends on data, it thrives on **structured, contextualized, and accessible knowledge.** This means: * Knowledge must be **findable and reusable**. * Teams must trust and understand **how knowledge flows** within the organization. * Decision-makers must know the **limits of what AI “knows.”** Without a KM strategy, AI becomes a “black box” surrounded by confusion, mistrust, and inefficiency. Let’s see how this plays out in the real world. ## Case Study 1: IBM Watson in Oncology – The Perils of Uncurated Knowledge ### Background: IBM’s Watson for Oncology was hailed as a game-changer. It was meant to digest thousands of medical journals and patient histories to help doctors recommend cancer treatments. ### What Went Wrong: * **Knowledge Inputs Were Incomplete or Inconsistent**: Watson’s suggestions were often based on limited or biased datasets, particularly from a single hospital( Memorial Sloan Kettering). * **No KM Governance**: There was no framework for validating which clinical knowledge would be prioritized, updated, or sunsetted. * **Doctors Didn’t Trust It**: Recommendations were sometimes irrelevant or contradicted clinical guidelines, undermining trust. ### KM Takeaway: Without **curated, contextualized, and expert-validated knowledge assets**, even the most powerful AI can make dangerous recommendations. A KM program could have ensured transparency in the data pipeline, regular reviews of training material, and clinician feedback loops to continuously refine outputs. ## Case Study 2: Shell – Combining KM and AI for Asset Integrity ### Background: Shell implemented AI to predict equipment failures across its oil and gas assets globally. But instead of starting with a tech-first approach, Shell focused on its** knowledge environment.** ### What Went Right: * **Integrated KM Strategy**: Shell had already invested in KM by mapping knowledge flows, codifying lessons learned, and standardizing reporting practices across global assets. * **Cross-Functional Knowledge Teams**: AI development involved engineers, data scientists, and KM practitioners working together. * **Knowledge as Training Fuel**: Historical maintenance logs, procedural checklists, and engineering insights were fed into the AI in structured formats. ### Outcome: Shell reported **30% improvements in predictive accuracy** and significant savings in downtime costs. ### KM Takeaway: Shell’s case proves that **knowledge isn’t just an input—it’s a strategic asset.** The AI was successful because the organization understood its knowledge landscape and embedded KM into the implementation process. ## Case Study 3: U.S. Department of Defense – AI for Logistics, Powered by Knowledge Engineering ### Background: The DoD has implemented AI in several domains, but logistics has seen some of the most measurable results. In one project, AI was used to forecast parts failures and optimize supply chain movements. ### What Worked: * **Ontology Development**: KM experts helped create taxonomies and knowledge maps of systems, parts, and supply relationships. * **Knowledge-Centric Change Management**: Users were trained not just in how to use the AI tools but in **how knowledge flows into and out of them.** * **KM Metrics**: The DoD tracked knowledge reuse, lessons captured, and decision accuracy alongside AI metrics. ### Outcome: The AI system reduced logistics planning time by over 40%, increased mission readiness, and improved confidence in predictive insights. ### KM Takeaway: By embedding AI into a **mature KM environment**, the DoD ensured its models were interpretable, trusted, and continuously updated. ## The KM Elements That Make AI Work To avoid failed implementations and maximize AI value, organizations must treat KM as **foundational**, not optional. Here’s what that looks like: ### 1. **Knowledge Strategy Alignment** AI must align with business-critical knowledge domains. A KM strategy identifies the knowledge that matters most—what needs to be captured, shared, and protected. **Example**: A bank implementing AI for fraud detection must ensure it has structured access to prior fraud case data, policies, and customer behavior profiles. ### 2. **Knowledge Mapping and Taxonomies** Before you train an AI, you must **know what you know**. Knowledge mapping identifies key sources, formats, and flows. **Without this**, AI will be fed fragmented, duplicated, or outdated data—leading to unreliable outputs. ### 3. **Content Governance** Who owns the knowledge? Who updates it? How often? Establishing governance ensures that AI is fed with **clean, current, and trustworthy knowledge**—and that decisions based on AI are auditable. ### 4. **Cultural Readiness** KM fosters a culture of **collaboration, transparency, and learning**. AI thrives in such cultures, where teams are open to machine-assisted insights and willing to contribute to knowledge improvement. ### 5. **Human-in-the-Loop Design** KM promotes **shared understanding** and bridges the AI-human divide. When users understand how AI makes decisions (thanks to a shared knowledge base), they’re more likely to trust and use it. ## How to Get Started: Embedding KM into Your AI Journey If you’re planning—or struggling through—an AI implementation, here’s a roadmap to integrate KM effectively: ### Phase 1: Assess * Conduct a KM Maturity Assessment. * Map critical knowledge assets and flows. * Identify knowledge gaps that will impact AI performance. ### Phase 2: Align * Align AI goals with the KM strategy. * Define success metrics for both AI and knowledge flow. * Involve KM professionals early in AI development. ### Phase 3: Build * Create taxonomies, metadata schemas, and knowledge repositories. * Implement knowledge curation workflows. * Ensure knowledge is machine-readable (structured data, tags, linked concepts). ### Phase 4: Govern & Sustain * Establish KM roles in AI operations (knowledge stewards, content owners). * Monitor knowledge quality and update cycles. * Use AI outputs to inform new knowledge creation (closed feedback loops). ## Conclusion: Smart AI Demands Smart KM AI will not replace people—it will replace **organizations that fail to manage their knowledge**. If you want your AI to deliver business value—whether through faster decisions, better customer service, or operational excellence—you must first build the knowledge infrastructure that fuels it. Knowledge Management is not the “back office” of your AI project. It’s the foundation. So before you invest another dollar in algorithms, ask yourself: > “Do we truly know what we know—and are we ready to teach it to our machines?” ## Final Thoughts: The Knoco International Approach At Knoco International, we’ve spent decades helping organizations across sectors design KM programs that unlock strategic value—especially when paired with emerging technologies like AI. We believe that the intersection of KM and AI is not just a technical opportunity, but a leadership imperative. If your AI initiative is stalling—or if you want to future-proof your implementation— let’s talk. Your knowledge is your edge. Let’s manage it wisely. ## In Conversation with… Cory Lee Cannon - Link: https://cannonco.net/in-conversation-with-cory-lee-cannon/ - Published: 2025-04-22T11:59:02-05:00 - Author: cannonco ![](https://cannonco.net/wp-content/uploads/2025/04/1745079042033-1024x576.png) **Ever wonder how other knowledge and information (K&IM) professionals work? How did they get into K&IM and to their current position? You will get a glimpse in this interview series with K&IM professionals, “In Conversation with…,” and perhaps even discover potential candidates for your career mentor.** _June Huang, K&IM professional, CILIP K&IM group committee member, London, UK_ [https://www.linkedin.com/pulse/conversation-cory-lee-cannon-june-huang-ssv5e](https://www.linkedin.com/pulse/conversation-cory-lee-cannon-june-huang-ssv5e)