<?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 Erim YAGCI on Medium]]></title>
        <description><![CDATA[Stories by Erim YAGCI on Medium]]></description>
        <link>https://medium.com/@erimyagci?source=rss-c89154d2a2fe------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*Sehme-dUS4-jMKumfAn8mw.jpeg</url>
            <title>Stories by Erim YAGCI on Medium</title>
            <link>https://medium.com/@erimyagci?source=rss-c89154d2a2fe------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 09:12:43 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@erimyagci/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 Data Architecture Reality Check]]></title>
            <link>https://medium.com/@erimyagci/the-data-architecture-reality-check-39a9d40da5c2?source=rss-c89154d2a2fe------2</link>
            <guid isPermaLink="false">https://medium.com/p/39a9d40da5c2</guid>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[data-architecture]]></category>
            <dc:creator><![CDATA[Erim YAGCI]]></dc:creator>
            <pubDate>Sat, 01 Aug 2026 11:03:34 GMT</pubDate>
            <atom:updated>2026-08-01T11:03:34.212Z</atom:updated>
            <content:encoded><![CDATA[<p>Before the BI or dashboard boom of the 2010s (led by Tableau, Power BI, and Looker), corporate data lived in fragmented, manual spreadsheets. Every department head brought their own conflicting Excel sheet to board meetings, leading to endless arguments about whose numbers were right. Dashboards introduced centralized, governed metrics that finally aligned Sales, Finance, and Marketing around a single source of truth. Dashboards were the gateway drug to data adoption in the corporate world. By putting visual analytics on screens in conference rooms and on executive laptops, they shifted organizational culture away from purely intuition based decisions (“gut feel”) toward evidence-based decision-making. But what was revolutionary in 2015 has become the latency bottleneck of 2026. yesterday’s command centers have become today’s passive reporting portals.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xIzBGphSXK2n7S927gWj_w.png" /><figcaption>Created with AI</figcaption></figure><p>Enterprise data architecture was sold on a simple promise: ingest enough TBs into a central repository, build enough dashboards, and competitive business intelligence would naturally follow. Today, that promise is facing a sharp reality check.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EDrO-n1KWcW071iexwuFFg.png" /><figcaption>Created with AI</figcaption></figure><p>The pattern is simple first we built centralized data warehouses. Then we needed data lakes. Then, we built data lakehouses. But the current target is about achieving three things:</p><ol><li><strong>Time-to-Value:</strong> How fast can a business domain turn raw data into a revenue-generating or cost-saving action? Every arrow in the development chain of dashboards introduces latency. Data engineering teams spend weeks writing bespoke ETL pipelines to land a new table. Analytics engineers spend days building custom dbt models. BI developers spend hours tweaking dashboard visualizations. Moving from <em>passive dashboards</em> (which require a human to read and react) to <em>embedded action triggers</em> (where data platforms call operational APIs to automatically update CRM records, trigger retention offers, or flag suspicious transactions).</li><li><strong>Unit Economics:</strong> Why does our cloud query bill double every time the business grows by 10%? FinOps is the operational framework and culture that aligns cloud, data, and AI infrastructure costs with actual business value. You have to enforce FinOps controls and implement strict resource monitors, auto-suspend rules, and department level cost tagging across data platforms on cloud like Snowflake, Databricks</li><li><strong>Agentic Readiness:</strong> Is our data foundation actually ready to power autonomous AI agents without hallucinating or leaking customer PII? The solution is The Semantic Layer as an interface for AI Agents. Please read my post about the <a href="https://medium.com/@erimyagci/the-missing-layer-in-enterprise-data-architecture-ontology-3ff272a7d050">Ontology</a> concept. To move from static BI to operational AI, our data layer must present clean, deterministic logic to AI engines. It may seem like a futuristic approach but, stop measuring your data team’s velocity by “dashboards built.” Start with identifying top value operational workflows and replace passive dashboard reporting with automated AI agent workflows supported by human sign off gates. Shift performance metrics toward “automated operational actions executed” and prioritize building a single, code governed semantic layer.</li></ol><h3>My Final Thoughts</h3><p>The era of building data infrastructure for the sake of larger volumes data is officially behind us. The winners in the modern enterprise landscape are not the companies with the biggest data lakes, but those who can move data seamlessly from storage to motion, from insights to autonomous execution, and from uncontrolled cloud spend to disciplined financial efficiency.</p><p>By focusing on open architectures, real-time execution, robust semantic layers, and operational MLOps, you will build a data platform that doesn’t just survive the next wave of AI innovation it will power it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=39a9d40da5c2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Missing Layer in Enterprise Data Architecture: Ontology]]></title>
            <link>https://medium.com/@erimyagci/the-missing-layer-in-enterprise-data-architecture-ontology-3ff272a7d050?source=rss-c89154d2a2fe------2</link>
            <guid isPermaLink="false">https://medium.com/p/3ff272a7d050</guid>
            <category><![CDATA[data]]></category>
            <category><![CDATA[ontology]]></category>
            <dc:creator><![CDATA[Erim YAGCI]]></dc:creator>
            <pubDate>Thu, 05 Mar 2026 18:50:58 GMT</pubDate>
            <atom:updated>2026-03-05T19:05:59.158Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>The Future of Competitive Advantage</strong></p><p>The next market leaders will not be the organizations with the most dashboards or the largest data lakes. Advantage is shifting away from simply owning data to unifying intelligence that turns data into outcomes.</p><p>By investing in an enterprise ontology, you elevate your organization above rigid tables and schemas. You create a living, real-time, structured view of your business that allows both your human teams and your AI agents to reason, decide, and act in the true language of your business.</p><p><strong>From Raw Data to Autonomous Intelligence: Why ‘Ontology’ is the Missing Link in Your Data Strategy</strong></p><p>Today’s winning organizations are not merely collecting more data; they are turning data into intelligence, and intelligence into action. We are witnessing a massive shift from traditional data platforms to unified intelligence platforms.</p><p>But there is a critical blocker in this evolution: AI and traditional data systems can read your data, but they do not understand your business. Nobody actually thinks in tables and schemas; leaders think in terms of planes, passengers, routes, or products. Currently, the true semantics of a business — how things connect, which policies apply, and what actions change results — live largely in people’s heads.</p><p>To bridge this gap and create massive value, modern data strategy is turning to a concept that has moved from artificial intelligence labs to the enterprise mainstream: the ontology.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jJ3RNqbQZEX6b7ZFrhtmnw.png" /></figure><p><strong>What Exactly is an Ontology?</strong></p><p>At its core, an ontology is a formal, explicit description of concepts in a domain of discourse. It acts as a blueprint or an enterprise vocabulary that defines your business objects and how they interact with each other.</p><p>Instead of just organizing data into columns and rows, an ontology maps out:</p><ul><li>Entities (Classes): These are the nouns of your business. They represent real-world or abstract concepts like a Customer, an Order, a Plane, or a Pilot.</li><li>Relationships (Verbs): This is how entities connect. Instead of simple numerical joins, relationships are defined by verbs. A Customer “places” an Order; a Pilot is “certified for” an Aircraft.</li><li>Properties (Slots): The specific facts describing the state of an entity, such as an aircraft’s fuel level, or a wine’s flavor and body.</li><li>Rules &amp; Actions: Triggers that initiate operations based on events, such as rebooking passengers if a connection time drops below 20 minutes.</li></ul><p><strong>Why Ontologies Unlock Unprecedented Value</strong></p><p>An ontology is not just a theoretical modeling tool; it is the semantic backbone for every team, application, and AI agent in your company. Here is how it drives value:</p><p>1. Creating a Shared, Cross-Domain Understanding A primary goal of developing an ontology is to share a common understanding of information among people and software agents. Every team typically has its own slice of the truth and its own definitions. By defining a unified enterprise vocabulary, you can seamlessly blend historical data (like data lakes) with real-time operational streams (like IoT telemetry) to achieve cross-domain consistency.</p><p>2. Supercharging AI and Autonomous Agents When you bind data using verbs and complex relationships, you give Large Language Models the exact context they need to understand what is truly going on in your business. An ontology provides AI with the same business awareness as your most experienced human experts, but at machine scale.</p><p>Furthermore, this semantic grounding enables Operations Agents — autonomous AI that doesn’t just answer questions, but runs operations. Because these agents understand your business’s rules and objectives through the ontology, they can monitor live signals, evaluate trade-offs, and take proactive, real-time actions across your operational systems.</p><p>3. Enabling Complex, Multi-Hop Reasoning Relational databases often silo information, complicating real-time analysis. Because ontologies function essentially as graph data, they are uniquely suited for complex relationships like supply chain optimization, IoT relationship traversal, and multi-hop AI reasoning.</p><p><strong>Microsoft’s Latest Launch: Fabric IQ</strong></p><p>IQ is a Microsoft Fabric workload designed to unify data across OneLake, providing a consistent semantic framework for analytics, AI agents, and applications. It organizes data according to business language, enhancing clarity and usability.</p><p>Fabric IQ (preview) enables a governed, self-service knowledge graph that links business entities, relationships, and business logic into a coherent framework for driving action across all enterprise data.</p><p>IQ is a collection of capabilities targeted to the common functionality of modeling an environment with unified language. The items grouped into the IQ workload include:</p><ul><li><a href="https://learn.microsoft.com/en-us/fabric/iq/ontology/overview">Ontology (preview)</a></li><li><a href="https://learn.microsoft.com/en-us/fabric/graph/overview">Graph (preview)</a></li><li><a href="https://learn.microsoft.com/en-us/fabric/data-science/concept-data-agent">Data agent (preview)</a></li><li><a href="https://learn.microsoft.com/en-us/fabric/real-time-intelligence/operations-agent">Operations agent (preview)</a></li><li><a href="https://learn.microsoft.com/en-us/fabric/data-warehouse/semantic-models">Power BI semantic models</a></li></ul><p><strong>Closing Thoughts: The Next Evolution of Data Platforms</strong></p><p>For many years, organizations invested heavily in building <strong>data lakes, warehouses, and dashboards</strong>. These platforms successfully centralized data and improved reporting, but the journey from <strong>data to real business value</strong> often remained incomplete. Data existed. Insights existed. Yet decisions and actions still required human interpretation across disconnected systems.</p><p>This is where <strong>ontology-driven architectures</strong> begin to redefine the future.</p><p>An ontology layer transforms data platforms from passive storage systems into <strong>operational intelligence systems</strong>. By structuring data around real-world business objects — customers, policies, claims, products, or risks — it allows analytics, AI models, and applications to interact with data in a way that reflects how the business actually operates.</p><p>We are already seeing early signals of this shift. Platforms such as <strong>Microsoft Fabric</strong> are evolving beyond traditional analytics infrastructure. Emerging capabilities like <strong>Fabric IQ</strong> point toward a future where AI agents can understand enterprise data models, reason over business entities, and generate insights or actions directly from the data platform itself.</p><p>In the coming years, several developments are likely to accelerate this transformation:</p><ul><li><strong>Semantic and ontology layers becoming core components of enterprise data platforms</strong></li><li><strong>AI agents interacting directly with business objects instead of raw tables</strong></li><li><strong>Data platforms evolving into decision and action platforms</strong></li><li><strong>Closer integration between data, AI models, and operational systems</strong></li></ul><p>In other words, the next generation of data platforms will not simply <strong>store and process data</strong> — they will <strong>understand it</strong>.</p><p>Organizations that embrace ontology-driven thinking today will be better positioned to build truly <strong>intelligent enterprises</strong>, where data flows naturally into insights, insights into decisions, and decisions into automated action.</p><p>The shift from <strong>data platforms to value platforms</strong> has already begun.<br> Ontology may well become the layer that finally makes it possible.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3ff272a7d050" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI is Disrupting the Shopping Process]]></title>
            <link>https://medium.com/@erimyagci/ai-is-disrupting-the-shopping-process-0ba326d3858f?source=rss-c89154d2a2fe------2</link>
            <guid isPermaLink="false">https://medium.com/p/0ba326d3858f</guid>
            <category><![CDATA[ecommerce]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[retail]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Erim YAGCI]]></dc:creator>
            <pubDate>Sun, 28 Dec 2025 13:45:50 GMT</pubDate>
            <atom:updated>2025-12-28T13:45:50.451Z</atom:updated>
            <content:encoded><![CDATA[<p>For most of people, AI means chatbots. We use tools like ChatGPT to answer questions, summarize documents, or help write an email. But behind this familiar interface, a much deeper and more important shift is happening: the rise of “intelligent agents” that don’t just talk, but act on our behalf.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y_9YdqVdbDIR232pRJgzfg.png" /></figure><p>These agents are software programs designed to perform complex tasks, from booking a multi-stop vacation to managing your inbox, all without direct human intervention. This post reveals five of the most surprising and impactful ways these AI agents are set to reshape our digital world, moving far beyond simple questions.</p><p><strong>AI Shopping Is Already Bigger Than You Think</strong></p><p>While it may seem like a futuristic concept, using AI for shopping has already achieved huge adoption. A survey by Shopify found that “around two-thirds of consumers in rich countries” and an overwhelming “five-sixths of those aged 18–24” plan to use AI to help them shop.</p><p>This isn’t just window shopping. A McKinsey study found that shopping advice is the second-most-common use for generative AI in America, and the consultancy projects that by 2030, a staggering $3trn-5trn of shopping worldwide will be conducted through such agents. This rapid behavioral shift is already creating strategic tension. E-commerce giant Amazon has blocked OpenAI’s crawlers and is suing AI firm Perplexity, aiming to keep customers on its own site. In stark contrast, its rival Walmart announced it would make its products available for purchase directly through ChatGPT, embracing agents as a new channel to reach customers.</p><p>A McKinsey study found that shopping advice is the second-most-common use for generative AI in America, and the consultancy projects that by 2030, a staggering $3trn-5trn of shopping worldwide will be conducted through such agents. This rapid behavioral shift is already creating strategic tension. E-commerce giant Amazon has blocked OpenAI’s crawlers and is suing AI firm Perplexity, aiming to keep customers on its own site. In stark contrast, its rival Walmart announced it would make its products available for purchase directly through ChatGPT, embracing agents as a new channel to reach customers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Jd6ID90ENXBAC4XzRQ2yeA.png" /><figcaption>Image created with NoteboolLM</figcaption></figure><p>Paradoxically, the rise of AI agents that dominate online browsing could make physical stores more important than ever. As AI agents become the primary interface for e-commerce, the websites of individual brands and retailers become less critical as a customer touchpoint.</p><p><strong>Marketers Are Preparing to Sell to Your AI, Not You</strong></p><p>The fundamental advertising model of the internet, built on capturing and monetizing human attention — a model that earns giants like Alphabet and Meta nearly half a trillion dollars a year — is on the verge of being upended. As agents begin making purchasing decisions on our behalf, marketers must shift their focus from persuading people to persuading their “digital proxy.”</p><p>This is giving rise to a new field: “generative-engine optimisation,” the successor to search-engine optimization (SEO). Brands are learning they must align their product data with how a user’s agent thinks.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kUFSyGiQydkzMh826GwgAQ.png" /><figcaption>Gemini AI Optimization</figcaption></figure><p><strong>In Conclusion</strong></p><p>We are rapidly moving beyond an internet where we must manually search, click, and browse. The core theme of this transformation is a shift from a “pull” internet, where we initiate every action, to a “push” internet, where intelligent agents act on our behalf, often without being prompted.</p><p>They will set up our meetings, flag relevant research, and handle countless small tasks, fundamentally changing our relationship with technology. This raises a final, critical question: As these agents become more capable, how much of our digital lives will we be willing to delegate, and what will it mean to be online when our agents are doing most of the work for us?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0ba326d3858f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Microsoft Fabric to unify data on a AI-Powered data platform]]></title>
            <link>https://medium.com/@erimyagci/microsoft-fabric-to-unify-data-on-a-ai-powered-data-platform-6be85d510c91?source=rss-c89154d2a2fe------2</link>
            <guid isPermaLink="false">https://medium.com/p/6be85d510c91</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data-engineering]]></category>
            <dc:creator><![CDATA[Erim YAGCI]]></dc:creator>
            <pubDate>Tue, 22 Jul 2025 06:07:05 GMT</pubDate>
            <atom:updated>2025-07-22T06:07:05.081Z</atom:updated>
            <content:encoded><![CDATA[<p>At its core, Microsoft Fabric is an end-to-end, unified analytics platform that brings together all the data and analytics tools that organizations need. Think of it as a single, comprehensive SaaS (Software as a Service) solution that integrates everything from data movement to data science, real-time analytics, and business intelligence.</p><p>The main features of the platform is about providing access to data through a unified lake — a single AI-ready source for all data needs.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*_bKQl4UnKcqbHQL9" /><figcaption>source: microsoft.com</figcaption></figure><h3>Why Data Leaders Should Pay Attention</h3><ul><li><strong>Simplified Architecture:</strong> Gone are the days of managing a dozen different vendors and complex integrations. Fabric provides a single product with a unified experience and architecture, dramatically reducing complexity.</li><li><strong>Breaking Down Silos:</strong> With all data and tools in one place, departmental and data silos naturally dissolve. Your data engineering, data science, and analytics teams can collaborate seamlessly on the same data.</li><li><strong>Enhanced Governance and Security:</strong> A unified platform means unified governance. Fabric makes it easier to manage security, privacy, and compliance across your entire data estate from a single control plane.</li><li><strong>Focus on Value, Not Integration:</strong> By freeing up resources previously spent on maintaining a patchwork of systems, leaders can focus their teams on what truly matters: deriving insights and driving business value from data.</li></ul><h3>Key Componets at a Glance</h3><ul><li><strong>Onelake: </strong>Connect data from any source into a single, multicloud data lake for the entire organization.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*Z9v2DYAqTSGnBHcV" /></figure><ul><li><strong>Data Factory:</strong> Connect and unify on-premises and cloud data sources with 180+ connectors for seamless integration. Enable efficient data synchronization with mirroring to ensure real-time consistency. Combine data transformation and self-service preparation for efficient, streamlined analytics. Orchestrate enterprise workflows with a seamless combination of Microsoft Fabric and Apache Airflow job.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*G6qOsfBRBSP6PM6p" /></figure><ul><li><strong>Analytics: </strong>Provide data engineers with authoring experiences that support data analysis and collaboration. Easily manage, secure, and analyze data of any size using SQL, no-code, or AI-driven experiences. Quickly create predictive AI models at scale and boost collaboration when deploying and managing machine learning models.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*h5-Yk1XBYRb1DsgQ" /></figure><ul><li><strong>Power BI: </strong>Create impactful reports and discover key insights from your data with easy-to-use tools and engaging visuals. Use natural language prompts to create reports, understand data, and summarize insights with Copilot in Power BI. Deliver data reports and insights directly within the Microsoft 365 apps you use every day.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*ajlD6qXWQ-woH4t5" /></figure><p>As a conclusion, Microsoft Fabric is more than a rebranding of familiar tools — it’s a bold step toward the data platform of the future. For organizations already invested in the Microsoft ecosystem, Fabric can dramatically simplify and enhance their data operations. Microsoft Fabric isn’t just an incremental improvement; it’s a transformative platform designed for the AI era. By unifying the entire analytics stack, it promises to simplify complexity, accelerate innovation, and unlock the full potential of your data</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6be85d510c91" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Future of Jobs: Insights from the World Economic Forum’s 2025 Report]]></title>
            <link>https://medium.com/@erimyagci/the-future-of-jobs-insights-from-the-world-economic-forums-2025-report-9c305f9a3200?source=rss-c89154d2a2fe------2</link>
            <guid isPermaLink="false">https://medium.com/p/9c305f9a3200</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[future-of-ai]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[jobs]]></category>
            <dc:creator><![CDATA[Erim YAGCI]]></dc:creator>
            <pubDate>Tue, 14 Jan 2025 13:16:36 GMT</pubDate>
            <atom:updated>2025-01-14T13:18:33.542Z</atom:updated>
            <content:encoded><![CDATA[<p>The World Economic Forum’s <em>Future of Jobs Report 2025</em> offers an in-depth analysis of the dynamic shifts reshaping industries, jobs, and skills. Here’s a summary of the key highlights and findings from this pivotal report, along with my reflections on the road ahead.</p><figure><img alt="Future of Jobs Report 2025 — World Economic Forum" src="https://cdn-images-1.medium.com/max/311/1*ZD0BTlsqQXDNnvjDxu4epw.jpeg" /><figcaption>Future of Jobs Report (Source: WEF)</figcaption></figure><p><strong>Key Highlights</strong></p><p><strong>1. Macrotrends Driving Transformation</strong></p><p>The report identifies five macrotrends that are reshaping the global labor market:</p><ul><li><strong>Technological Change:</strong> Advancements in artificial intelligence (AI), robotics, and digital access are expected to transform 60% of businesses by 2030.</li><li><strong>Green Transition:</strong> Climate change mitigation and adaptation are driving growth in renewable energy roles and environmental engineering.</li><li><strong>Economic Uncertainty:</strong> Rising costs of living and slower economic growth are transforming workforce strategies and reshaping job demands.</li><li><strong>Geoeconomic Fragmentation:</strong> Increased trade restrictions and geopolitical tensions are fueling demand for cybersecurity and resilience skills.</li><li><strong>Demographic Shifts:</strong> Aging populations in high-income economies and expanding working-age populations in low-income countries are redefining talent availability and workforce priorities.</li></ul><p><strong>Jobs Outlook:</strong></p><ul><li><strong>Fastest-Growing Roles</strong>: AI and Machine Learning Specialists, Big Data Specialists, and Renewable Energy Engineers lead the growth trajectory.</li><li><strong>Declining Roles</strong>: Clerical jobs, such as Data Entry Clerks and Bank Tellers, are expected to face significant displacement due to automation and digitalization.</li><li><strong>Net Employment Growth</strong>: By 2030, 170 million jobs are projected to be created, offset by 92 million job displacements, resulting in a net growth of 7% or 78 million jobs globally.</li></ul><p><strong>Skills Outlook:</strong></p><ul><li><strong>Evolving Needs</strong>: Analytical thinking, resilience, and adaptability remain top skills, complemented by emerging priorities like AI literacy and environmental stewardship.</li><li><strong>Reskilling Imperative</strong>: 59% of the workforce will require upskilling by 2030 to remain relevant, with employers emphasizing in-house training and redeployment strategies.</li></ul><p><strong>Geoeconomic and Societal Challenges:</strong></p><ul><li>Geopolitical tensions and trade restrictions will influence global supply chains and business models, driving demand for cybersecurity and leadership skills.</li><li>Diversity, equity, and inclusion initiatives have gained prominence, with 83% of employers adopting such strategies.</li></ul><h3><strong>Deep Dive into Skills Outlook: Reshaping Workforce Capabilities</strong></h3><p>As we approach 2030, the demand for skills is undergoing a seismic shift, driven by rapid technological advancements, economic transformations, and societal needs. Here are the key insights and trends:</p><p><strong>1. Top Skills in Demand</strong></p><p>The report identifies the following as the most sought-after skills by 2030:</p><ul><li><strong>Analytical Thinking</strong>: Continues to dominate as the most essential skill, with 70% of employers highlighting its importance. Analytical thinking encompasses problem-solving, decision-making, and the ability to interpret data effectively.</li><li><strong>Resilience, Flexibility, and Agility</strong>: With rising uncertainties in global markets, employees who can adapt to change, recover quickly from setbacks, and pivot in dynamic situations are highly valued.</li><li><strong>Leadership and Social Influence</strong>: As workplaces adopt more automation, human-centered leadership skills are vital to guide teams through change and foster collaboration.</li><li><strong>Technology-Related Skills</strong>:</li><li><strong>AI and Big Data Proficiency</strong>: Employers foresee growing demand for skills in artificial intelligence, machine learning, and data analysis.</li><li><strong>Networks and Cybersecurity</strong>: Protecting digital assets has become a strategic priority, driving demand for expertise in these areas.</li><li><strong>Technological Literacy</strong>: Beyond technical roles, all employees are expected to have a baseline understanding of emerging technologies.</li><li><strong>Creative Thinking and Lifelong Learning</strong>: Innovation and a commitment to continuous skill development are seen as critical to staying competitive in the workforce.</li></ul><p><strong>2. Emerging Skills</strong></p><p>The report also highlights newer priorities in skills development, reflecting broader shifts in the global economy:</p><ul><li><strong>Environmental Stewardship</strong>: With the green transition accelerating, skills related to sustainable practices and climate adaptation are now among the top 10 growing competencies.</li><li><strong>Curiosity and Lifelong Learning</strong>: Employers are increasingly valuing an inherent drive for knowledge, especially in a landscape where job roles and tools evolve rapidly.</li></ul><p><strong>3. Skills Becoming Obsolete</strong></p><p>Some traditional skills are witnessing a decline in relevance due to automation and changing job demands:</p><ul><li><strong>Manual Dexterity, Endurance, and Precision</strong>: Automation and robotics are replacing repetitive, physical tasks.</li><li><strong>Clerical and Administrative Skills</strong>: Roles such as data entry and routine documentation are becoming automated, reducing the need for these skills.</li></ul><p><strong>4. Workforce Reskilling Imperative</strong></p><p>The scale of upskilling and reskilling required is monumental:</p><ul><li><strong>39% Skill Transformation</strong>: By 2030, nearly two-fifths of workers’ current skill sets will need to be updated or replaced.</li><li><strong>Upskilling Initiatives</strong>:</li><li><strong>In-role Upskilling</strong>: 29% of the workforce can be reskilled in their current roles.</li><li><strong>Redeployment Upskilling</strong>: 19% can be reskilled and moved into new roles within the organization.</li><li><strong>Unaddressed Skill Gaps</strong>: Alarmingly, 11% of workers may not receive the reskilling required, leaving them vulnerable to displacement.</li></ul><p><strong>5. Barriers to Skill Development</strong></p><p>The report identifies significant challenges organizations face in addressing skill gaps:</p><ul><li><strong>Costs of Reskilling</strong>: Limited budgets for training programs.</li><li><strong>Time Constraints</strong>: Difficulty in balancing operational needs with training schedules.</li><li><strong>Talent Mismatch</strong>: Insufficient availability of trainers or curriculum tailored to new job requirements.</li></ul><p><strong>6. Strategic Solutions</strong></p><p>Organizations are implementing several strategies to address these challenges:</p><ul><li><strong>Internal Training Programs</strong>: Leveraging in-house resources for targeted skill development.</li><li><strong>Diversity, Equity, and Inclusion (DEI) Initiatives</strong>: Broadening talent pools by focusing on underrepresented groups.</li><li><strong>Public-Private Partnerships</strong>: Collaborating with educational institutions and governments to co-develop training programs.</li><li><strong>AI-Driven Learning Platforms</strong>: Utilizing generative AI tools to deliver personalized, scalable training modules.</li></ul><p><strong>Closing Reflections</strong></p><p>The Future of Jobs Report 2025 reminds us that adaptability and foresight are key to thriving amid transformation. It underscores the role of technology as both an enabler and disruptor and calls for investments in upskilling and sustainable practices to bridge emerging skill gaps and foster inclusive growth.</p><p>As businesses and leaders, we must embrace these changes proactively — aligning workforce strategies with the dynamic needs of the future. Whether through adopting AI-driven innovations, championing green transitions, or addressing societal disparities, we have a unique opportunity to reshape the world of work for the better.</p><p><strong>Source:</strong> <em>Future of Jobs Report 2025</em>, World Economic Forum.</p><p>Available at: <a href="http://www.weforum.org/reports/the-future-of-jobs-report-2025.">www.weforum.org/reports/the-future-of-jobs-report-2025</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9c305f9a3200" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Adapt or Vanish: The Critical Role of Organizational Agility in Today’s Dynamic Business Landscape]]></title>
            <link>https://medium.com/@erimyagci/adapt-or-vanish-the-critical-role-of-organizational-agility-in-todays-dynamic-business-landscape-13da0c0178e6?source=rss-c89154d2a2fe------2</link>
            <guid isPermaLink="false">https://medium.com/p/13da0c0178e6</guid>
            <category><![CDATA[organizational-agility]]></category>
            <dc:creator><![CDATA[Erim YAGCI]]></dc:creator>
            <pubDate>Wed, 13 Nov 2024 06:35:03 GMT</pubDate>
            <atom:updated>2024-11-13T06:35:03.980Z</atom:updated>
            <content:encoded><![CDATA[<p>In an era marked by rapid technological advancements, shifting economic tides, and evolving consumer expectations, the business landscape is more dynamic than ever before. Companies that once dominated their industries now struggle to stay afloat, while agile startups disrupt markets overnight. This article explores the imperative of organizational agility in navigating the changing business and economic environments and offers strategies to avoid the pitfalls that have led to the downfall of once-thriving companies.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/225/0*D6iuvfZuSMo6LZy5" /><figcaption>Woolly mammoth</figcaption></figure><p><strong>The Changing Environment</strong></p><p>The modern business landscape is a complex and constantly shifting terrain, shaped by an array of forces that extend beyond the boardroom. Rapid technological advancements, economic uncertainties, shifting societal values, and environmental pressures are collectively redefining how organizations operate. To thrive in this dynamic environment, companies must become increasingly agile and proactive.</p><p>The digital revolution has accelerated with unprecedented speed, transforming how businesses operate and interact with consumers. Artificial intelligence, machine learning, and automation have moved from futuristic concepts to core elements of daily operations across industries. These technologies enable companies to analyze vast amounts of data, personalize customer interactions, and streamline their supply chains. Yet, as quickly as some companies adopt these innovations, others struggle to keep pace, unable to pivot as new tools emerge. The result is a widening gap between industry leaders who adapt swiftly and those unable to break free from outdated models.</p><p>Compounding these technological shifts are the economic instabilities that regularly disrupt markets worldwide. Rising inflation, currency fluctuations, and unpredictable recessions have become recurrent challenges, requiring companies to remain resilient and flexible in their pricing, sourcing, and production strategies. Businesses that cannot adapt to economic pressures face tough decisions on downsizing, cost-cutting, or restructuring, risking long-term stability. Meanwhile, regulatory changes, imposed to keep pace with technological and market developments, add further layers of complexity, creating compliance demands that can stifle agility if not carefully managed.</p><p>Alongside economic and technological transformations, the social environment is undergoing profound shifts. Consumers today expect more from the brands they support, seeking companies that align with their personal values around sustainability, diversity, and corporate responsibility. Ethical consumerism has taken hold, and businesses that fail to embody transparent, responsible practices increasingly face backlash or customer abandonment. In parallel, employees also demand more from employers, valuing inclusive and purpose-driven workplaces as much as job security or benefits. This “social contract” shift requires companies to rethink not just their customer relationships but their organizational cultures and employer brand.</p><p>Environmental challenges further underscore the urgency for change. As the planet grapples with climate change, resource scarcity, and increasing regulation around carbon emissions, companies are pressed to adopt sustainable practices. Adapting to these pressures is no longer a matter of public relations; it’s a business imperative. For some industries, it means rethinking supply chains to reduce carbon footprints; for others, it means designing products with sustainable materials or adopting renewable energy sources to stay competitive and compliant with emerging environmental standards.</p><p>Together, these technological, economic, social, and environmental changes form a landscape that demands responsiveness, innovation, and adaptability. Traditional models of stability and long-term predictability are fading, replaced by a paradigm that values flexibility and foresight. To survive — and thrive — companies must not only recognize these interwoven challenges but also embrace a mindset of continual evolution. Organizations willing to adapt swiftly to this multifaceted landscape will lead the way, while those clinging to old ways of thinking risk obsolescence in a world that never stops evolving.</p><p><strong>Why Companies Fail to Adapt</strong></p><p><strong>Resistance to Change:</strong> Organizational inertia and comfort with the status quo hinder adaptation. <strong>Lack of Innovation:</strong> Failing to invest in R&amp;D can make products and services obsolete. <strong>Poor Leadership Vision:</strong> Without forward-thinking leadership, companies cannot navigate future challenges.</p><p><strong>Creating Awareness About the Need for Agility</strong></p><ol><li><strong>Leadership and Culture</strong></li></ol><ul><li><strong>Championing Change:</strong> Leaders must advocate for agility and model adaptive behaviors.</li><li><strong>Cultivating a Learning Environment:</strong> Encourage continuous learning and openness to new ideas.</li></ul><p><strong>2. Stakeholder Engagement</strong></p><ul><li><strong>Transparent Communication:</strong> Keep employees, investors, and customers informed about changes and strategies.</li><li><strong>Feedback Mechanisms:</strong> Implement systems to gather input from all levels of the organization.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/272/0*c0ka5n12E5gznJY8" /><figcaption>Ant Colony</figcaption></figure><p><strong>Strategies to Avoid Organizational Decline</strong></p><ol><li><strong>Adopt Flexible Structures</strong></li></ol><ul><li><strong>Agile Teams:</strong> Create cross-functional teams that can quickly respond to changes.</li><li><strong>Decentralized Decision-Making:</strong> Empower lower levels of the organization to make swift decisions.</li></ul><p><strong>2. Embrace Technological Innovation</strong></p><ul><li><strong>Invest in Emerging Technologies:</strong> Stay ahead by adopting AI, machine learning, and other innovations.</li><li><strong>Digital Customer Engagement:</strong> Utilize digital platforms to enhance customer experience.</li></ul><p><strong>3. Focus on Customer-Centricity</strong></p><ul><li><strong>Personalized Services:</strong> Use data analytics to tailor offerings to individual customer needs.</li><li><strong>Customer Feedback Loops:</strong> Actively seek and act on customer input to improve products and services.</li></ul><p><strong>4. Strategic Planning and Risk Management</strong></p><ul><li><strong>Scenario Planning:</strong> Prepare for multiple future scenarios to remain resilient.</li><li><strong>Diversification:</strong> Spread risk by diversifying products, services, and markets.</li></ul><p><strong>Conclusion</strong></p><p>The fall of once-dominant companies serves as a stark reminder that success is not guaranteed in today’s fast-paced world. Organizational agility is no longer a luxury but a necessity. By recognizing the signs of a changing environment and proactively adapting, companies can not only survive but thrive amid uncertainty. Leaders must act now to embed agility into the very fabric of their organizations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=13da0c0178e6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[OpenAI Sept. 2024 Update— New Character unlocked: Reasoning Models]]></title>
            <link>https://medium.com/@erimyagci/openai-sept-2024-update-new-character-unlocked-reasoning-models-a1f8d1b09931?source=rss-c89154d2a2fe------2</link>
            <guid isPermaLink="false">https://medium.com/p/a1f8d1b09931</guid>
            <category><![CDATA[generative-ai-tools]]></category>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Erim YAGCI]]></dc:creator>
            <pubDate>Fri, 13 Sep 2024 08:38:00 GMT</pubDate>
            <atom:updated>2024-09-13T08:38:00.896Z</atom:updated>
            <content:encoded><![CDATA[<p>Today 12.09.2024 Open AI ChatGPT Plus users have received an email informing new series of models. On this e-mail;</p><p>Users are informed that they can try the o1 beta in ChatGPT Plus. OpenAI is rolling out two models:</p><ul><li>Larger model, o1-preview, which has strong reasoning capabilities and broad world knowledge.</li><li>Smaller model, o1-mini, which is faster and is competitive with o1-preview at coding tasks</li></ul><p>Both models are accesible on Chat GPT web and mobile. <strong>ChatGPT Plus and Team</strong> users will be able to access o1 models in ChatGPT starting 12.09.2024. Both o1-preview and o1-mini can be selected manually in the model picker, and at launch, weekly rate limits will be 30 messages for o1-preview and 50 for o1-mini. Open AI is working to increase those rates and enable ChatGPT to automatically choose the right model for a given prompt.</p><p>Let’s deep dive into these two models.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*2XjXYgnzJJdNZHh1.jpeg" /><figcaption>ChatGPT Model View</figcaption></figure><h3><strong>Open AI o1-Preview (Uses Advanced Reasoning)</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*obvbRWtj84Ck_bFg.png" /><figcaption>Open AI o1-Preview</figcaption></figure><p>Open AI have developed a new series of AI models designed to spend more time thinking before they respond. They can reason through complex tasks and solve harder problems than previous models in science, coding, and math. This is a preview and we expect regular updates and improvements. Through training, these new generation models learn to refine their thinking process, try different strategies, and recognize their mistakes.</p><p>Enhanced reasoning capabilities may be particularly useful if you’re tackling complex problems in science, coding, math, and similar fields. For example, o1 can be used by healthcare researchers to annotate cell sequencing data, by physicists to generate complicated mathematical formulas needed for quantum optics, and by developers in all fields to build and execute multi-step workflows.</p><p>In a qualifying exam for the International Mathematics Olympiad (IMO), GPT-4o correctly solved only 13% of problems, while the reasoning model scored 83%. Their coding abilities were evaluated in contests and reached the 89th percentile in Codeforces competitions.</p><h3>Open AI o1-Mini (Advancing cost-efficient reasoning)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*jE-f5ejw_rR0b6e8" /><figcaption>OpenAI o1-mini</figcaption></figure><p>Large language models such as o1 are pre-trained on vast text datasets. While these high-capacity models have broad world knowledge, they can be expensive and slow for real-world applications. In contrast, o1-mini is a smaller model optimized for STEM reasoning during pretraining. After training with the same high-compute reinforcement learning (RL) pipeline as o1, o1-mini achieves comparable performance on many useful reasoning tasks, while being significantly more cost efficient.</p><p><strong>Mathematics: </strong>In the high school AIME math competition, o1-mini (70.0%) is competitive with o1 (74.4%)–while being significantly cheaper–and outperforms o1-preview (44.6%). o1-mini’s score (about 11/15 questions) places it in approximately the top 500 US high-school students.</p><p><strong>Coding:</strong> On the Codeforces competition website, o1-mini achieves 1650 Elo, which is again competitive with o1 (1673) and higher than o1-preview (1258). This Elo score puts the model at approximately the 86th percentile of programmers who compete on the Codeforces platform. o1-mini also performs well on the HumanEval coding benchmark and high-school level cybersecurity capture the flag challenges (CTFs).</p><p><strong>STEM:</strong> On some academic benchmarks requiring reasoning, such as GPQA (science) and MATH-500, o1-mini outperforms GPT-4o. o1-mini does not perform as well as GPT-4o on tasks such as MMLU and lags behind o1-preview on GPQA due to its lack of broad world knowledge.</p><h3><strong>Updated Model Lineup</strong></h3><p>To better understand o1-preview and o1-mini’s place in Open AI’s current model lineup, here’s a quick overview of the key models powering Azure OpenAI Service:</p><ul><li><strong>o1-preview: </strong>Focused on advanced reasoning and solving complex problems, including math and science tasks. Ideal for applications that require deep contextual understanding and agentic workflows.</li><li><strong>o1-mini:</strong> Smaller and faster, and 80% cheaper than o1-preview, performs well at code generation and small context operations.</li><li>GPT-4o: A versatile, multimodal model that excels in both text and image processing, with superior performance in non-English languages and vision tasks. Suitable for applications needing enhanced accuracy and multilingual capabilities. The model also features JSON Structured Outputs for consistent, well-defined data formats, reducing post-processing needs and improving application efficiency. Designed for real-time applications that require fast, reliable text responses at minimal cost.</li><li>GPT-4o Mini: A smaller, cost-effective version of GPT-4o, optimized for environments with limited resources or high cost constraints. Retains text and image processing capabilities, making it ideal for lightweight applications.</li><li>DALL-E: Generates images from text prompts with safety, ideal for creative content and marketing.</li><li>Whisper: Transcribes and translates speech to text, suitable for real-time transcription and multilingual communication.</li></ul><h3>What’s next</h3><p>This is an early preview of these reasoning models in ChatGPT and the API. In addition to model updates, it is expected addition of browsing, file and image uploading, and other features to make them more useful to everyone.</p><p><strong><em>Sources: </em></strong><em>Open AI Newsletter, Open AI Website, Microsoft Web Site</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a1f8d1b09931" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[OpenAI Spring Update - What comes with GPT-4o]]></title>
            <link>https://medium.com/@erimyagci/openai-spring-update-what-comes-with-gpt-4o-bcfab07eafb1?source=rss-c89154d2a2fe------2</link>
            <guid isPermaLink="false">https://medium.com/p/bcfab07eafb1</guid>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[chatgpt4]]></category>
            <dc:creator><![CDATA[Erim YAGCI]]></dc:creator>
            <pubDate>Mon, 13 May 2024 18:36:08 GMT</pubDate>
            <atom:updated>2024-05-13T20:47:23.968Z</atom:updated>
            <content:encoded><![CDATA[<p>OpenAI has unveiled GPT-4o, a new AI model that combines text, vision, and audio. The use cases shocked the world.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6OQGOXER6dWflBBy1VQ-7g.jpeg" /><figcaption>OpenAI’s new GPT-4o model can help with math homework. Credit: OpenAI</figcaption></figure><p>OpenAI’s previous leading model, an enhanced version of GPT-4 — was trained a combination of images and text, and could analyze images and text to accomplish tasks like extracting text from images or even describing the content of those images.</p><p>GPT-4o adds speech to the mix. In addition GPT-4o, OpenAI’s new flagship model can reason across audio, vision, and text in real time.</p><p>GPT-4o will be available for free to all ChatGPT users. It is also available in the API, and is half the price and twice as fast as GPT -4 Turbo. The “o” in the name stands for “omni,” referencing its combined modalities in one model.</p><p>It can also sense emotions and tones and express its own emotions and tones, making it sound extremely dramatic or robotic. It can even sing.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FDQacCB9tDaw%3Fstart%3D4&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DDQacCB9tDaw&amp;image=http%3A%2F%2Fi.ytimg.com%2Fvi%2FDQacCB9tDaw%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/2712d5fc824c6bf5af50dcebdd9e8cd5/href">https://medium.com/media/2712d5fc824c6bf5af50dcebdd9e8cd5/href</a></iframe><p>By highlighting code on the screen, ChatGPT with GPT-4o can process and understand what the code is and help improve it. In addition, given a photo ChatGPT can now quickly answer related question like, “What brand of shirt is this person wearing?”</p><p>Another new feature is that; users can ask ChatGPT a question and interrupt ChatGPT while it’s answering. The model delivers “real time” responsiveness.</p><p>Here are some mind blowing use cases of GPT-4o:</p><ol><li>Realtime Translation with GPT-4o</li></ol><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FWzUnEfiIqP4%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DWzUnEfiIqP4&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FWzUnEfiIqP4%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/90d5c48374ae8eaae7f642ea0d322e11/href">https://medium.com/media/90d5c48374ae8eaae7f642ea0d322e11/href</a></iframe><p>2. Tutoring with GPT-4o</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F_nSmkyDNulk%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D_nSmkyDNulk&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F_nSmkyDNulk%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/47c59bba060f984a50ff398b5c42d1a1/href">https://medium.com/media/47c59bba060f984a50ff398b5c42d1a1/href</a></iframe><p>3 — Be My Eyes Accessibility with GPT-4o</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FKwNUJ69RbwY%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DKwNUJ69RbwY&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FKwNUJ69RbwY%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/1703437dd404d0a275f31e567692e315/href">https://medium.com/media/1703437dd404d0a275f31e567692e315/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bcfab07eafb1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Data to AI on Google Cloud Platform]]></title>
            <link>https://medium.com/@erimyagci/data-to-ai-on-google-cloud-platform-346c4ebc3aa6?source=rss-c89154d2a2fe------2</link>
            <guid isPermaLink="false">https://medium.com/p/346c4ebc3aa6</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[mlops]]></category>
            <category><![CDATA[gcp]]></category>
            <category><![CDATA[google-cloud-platform]]></category>
            <dc:creator><![CDATA[Erim YAGCI]]></dc:creator>
            <pubDate>Fri, 19 Apr 2024 18:34:03 GMT</pubDate>
            <atom:updated>2024-04-19T18:39:30.385Z</atom:updated>
            <content:encoded><![CDATA[<p>After 3 years publishing <a href="https://medium.com/@erimyagci/modern-data-warehouse-on-google-cloud-platform-d91cc0b3b7c">Modern Data Warehouse on Google Cloud Platform</a>, It was time to talk about AI tools on GCP. Cloud AI is deployment of artificial intelligence models and services in the cloud, is becoming increasingly important for businesses across various industries. By implementing cloud AI, companies can focus mostly on their core business activities rather than on the complexities of managing AI infrastructure. This shift in focus can lead to better innovation and improved business outcomes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UQOKmxywp10BZSu0p15-ow.jpeg" /></figure><p>As an introduction; GCP is a suite of cloud computing services offered by Google. It runs on the same infrastructure that Google uses internally for its end-user products, such as Google Search, Gmail, file storage, and YouTube.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*s4ub62ocj7ksAHNMLlhc2g.png" /><figcaption>Google Cloud Platform</figcaption></figure><p>“Data to AI” is the process or methodologies involved in transforming raw data into actionable insights using AI technologies. This process is crucial for today’s businesses and organizations that want to leverage their data to generate business value on their businesses.</p><p>As you can see below, there are different stages for the data to go through when building a data to AI solution, including data engineering, data analysis, feature engineering, model training, model deployment, and MLOps. Let’s take a closer look at each of these stages and the Google Cloud products you can use to enable each stage.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*vPSxxYxT-p_0_a-_HHMYlg.png" /></figure><p>These are the list of Data to AI Steps and Google Cloud’s Data and Analytics products, must have for your GCP on AI Cheat Sheet.</p><p><strong>Data Engineering:</strong> <em>Dataflow</em>, a unified stream and batch data processing solution that’s serverless, fast, and cost-effective. My favorite tool for analytics projects <em>BigQuery</em>, a fully managed, petabyte-scale data warehouse that makes it easy to analyze large datasets of structured, semi-structured, and unstructured data</p><p><strong>Data Analysis: </strong>BigQuery has a built-in editor that allows you to run and execute SQL directly in BigQuery, but you can also consider using Looker, a business intelligence solution for analyzing, exploring, and visualizing your data, or Vertex AI Workbench, a Jupyter notebook-based development environment where you can access and explore your data from within a Jupyter notebook using BigQuery and Cloud Storage integrations.</p><p><strong>Model Training: </strong>Automated machine learning; Create and train models on image, tabular, text, and video datasets without writing code. BigQuery ML: Train models directly in BigQuery using SQL. Manage, orchestrate, and register it directly to Vertex AI Model Registry. Custom training: Bring your own code and train models using OSS frameworks (scikit-learn, TensorFlow, PyTorch, XGBoost, etc.), or bring your own container and train using Vertex AI Training.</p><p><strong>Model Deployment:</strong> Online predictions are synchronous requests made to a model endpoint. Use online predictions when you are making requests in response to application input or in situations that require timely inference. Batch predictions are asynchronous requests. You request batch predictions directly from the model resource without needing to deploy the model to an endpoint. Use batch predictions when you don’t require an immediate response and want to process accumulated data by using a single request.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*_d-vuJAAo8oHNzxfkQDSYQ.png" /></figure><p><strong>MLOps:</strong> Vertex AI provides a host of solutions to monitor and govern your models, all of which help drive successful and responsible AI deployment. In particular, Vertex AI Model Monitoring is a solution with anomaly and drift detection, reporting, alerting, and recommendation capabilities to help you ensure:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*My8dTb-YRpM7c_6K7CRsqQ.png" /></figure><p>In conclusion, the journey from data to AI Value driven solution development multi-faceted. As we’ve explored, each stage from data engineering to model deployment is important and offers its own set of challenges. GCP tools and services provide everything you need to effectively navigate this journey.</p><p>Embracing these technologies not only accelerates innovation but also empowers your organization to harness the true power of your data, making smarter decisions and staying competitive in a rapidly evolving digital landscape. As maturity of AI development continues to grow, the integration MLOps become increasingly essential.</p><p>Sources:</p><p><a href="https://www.googlecloudcommunity.com/"><em>https://www.googlecloudcommunity.com/</em></a></p><p><a href="https://cloud.google.com/vertex-ai/?hl=en"><em>Innovate faster with enterprise-ready AI, enhanced by Gemini models</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=346c4ebc3aa6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Driving Safety vs Physical Buttons]]></title>
            <link>https://medium.com/@erimyagci/driving-safety-vs-physical-buttons-bdc4d241a200?source=rss-c89154d2a2fe------2</link>
            <guid isPermaLink="false">https://medium.com/p/bdc4d241a200</guid>
            <category><![CDATA[insurance]]></category>
            <category><![CDATA[data-analytics]]></category>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[insurtech]]></category>
            <dc:creator><![CDATA[Erim YAGCI]]></dc:creator>
            <pubDate>Tue, 05 Mar 2024 16:59:39 GMT</pubDate>
            <atom:updated>2024-03-05T16:59:39.599Z</atom:updated>
            <content:encoded><![CDATA[<h3>Cars will need fewer screens and more buttons to earn 5-star safety rating in Europe</h3><p>The invention of the motor car was a major leap forward in human history; since then, it has propelled mankind to greater heights in the last 200 years.</p><p>The first recorded motor accident in history, as it happens, was in 1891 in Ohio, USA — five years after the invention of the car — and the first death caused by a motor accident in 1899. Since then, countless people have experienced motor accidents in various degrees of severity.</p><p>According to <a href="https://www.asirt.org/safe-travel/road-safety-facts/">annual global road crash statistics</a>, 1.35 million people lose their lives in road accidents every year — an average of 3,700 deaths every day. A further 20–50 million also suffer non-fatal injuries due to road accidents.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*ibMuNpE5bSpjkl392kIM2g.png" /><figcaption>Car Dashboard</figcaption></figure><p>In the last few decades, these trends have pushed governments across the world to put their best foot forward to increase road safety, and that resulted in significant decreases in the number of accidents.</p><p>In recent years, however, road accidents have, once more, seen a gradual increase. This raises important questions about the effectiveness of traditional road safety measures, creating a real need for more data analystics methods to improve road safety.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/360/1*9M8TN8uNVa15tk2i72odjg.png" /></figure><p>The European New Car Assessment Programme (Euro NCAP) is a European voluntary car safety performance assessment programme (i.e. a New Car Assessment Program) based in Leuven, Belgium. Formed in 1996, the first results were released in February 1997.</p><p>The Euro NCAP’s safety guidelines aren’t a legal requirement, however, car makers take safety ratings pretty seriously, so any risk of points being docked during such assessments is likely to be taken into consideration.</p><p><em>“The overuse of touchscreens is an industry-wide problem, with almost every vehicle-maker moving key controls onto central touchscreens, obliging drivers to take their eyes off the road and raising the risk of distraction crashes,” </em>said Matthew Avery, director of strategic development at Euro NCAP</p><p>He further added, <em>“New Euro NCAP tests due in 2026 will encourage manufacturers to use separate, physical controls for basic functions in an intuitive manner, limiting eyes-off-road time and therefore promoting safer driving.”</em></p><p>Some manufacturers like Tesla and Volkswagen have gained a reputation for placing basic vehicle controls behind touch-sensitive interfaces.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tpIns-gxtJHbxNp6DRDjAg.jpeg" /><figcaption>No Button Dashoard</figcaption></figure><p>This new regulation will have a huge impact on automotive manufacturers. Although I find the regulation very appropriate, I think it should be explained what data and insights it is based on. To which user segment the data relies on, Digital Natives playing with tablets since they are born, are they distracted with larger screens or phsyical buttons?</p><p><em>Sources: </em><a href="https://www.theverge.com/2024/3/5/24091043/euro-ncap-safety-rating-europe-2026-touchscreen-buttons-dials"><em>https://www.theverge.com/2024/3/5/24091043/euro-ncap-safety-rating-europe-2026-touchscreen-buttons-dials</em></a></p><p><a href="https://www.team-bhp.com/news/euro-ncap-wants-carmakers-bring-back-physical-buttons"><em>https://www.team-bhp.com/news/euro-ncap-wants-carmakers-bring-back-physical-buttons</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bdc4d241a200" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>