<?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 Anil Verma on Medium]]></title>
        <description><![CDATA[Stories by Anil Verma on Medium]]></description>
        <link>https://medium.com/@anilvermaspeaks?source=rss-41d4d241429------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*Kyqt2B_IIF1IU0rxhEjBgg.jpeg</url>
            <title>Stories by Anil Verma on Medium</title>
            <link>https://medium.com/@anilvermaspeaks?source=rss-41d4d241429------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 04:13:41 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@anilvermaspeaks/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[What is AI RAG?]]></title>
            <link>https://anilvermaspeaks.medium.com/what-is-ai-rag-4d6a371a30d9?source=rss-41d4d241429------2</link>
            <guid isPermaLink="false">https://medium.com/p/4d6a371a30d9</guid>
            <category><![CDATA[claude]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[rags]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[openai]]></category>
            <dc:creator><![CDATA[Anil Verma]]></dc:creator>
            <pubDate>Sat, 11 Oct 2025 12:34:35 GMT</pubDate>
            <atom:updated>2025-10-11T12:36:20.184Z</atom:updated>
            <content:encoded><![CDATA[<p>RAG doesn’t just rely on the information it has been trained on. Instead, it can dynamically retrieve relevant documents or data from an external knowledge base (like a database, the web, or a collection of documents) and then generate answers or content based on that newly fetched information.</p><p>This approach helps AI models become more accurate and contextually aware, reducing the limitations of traditional, closed-set models.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ymyDsg9pzSCxNHHUaKjKvQ.jpeg" /></figure><h4><strong>Pillars of RAG: Retrieval and Generation</strong></h4><p><strong>Retrieval</strong>:<br>In the retrieval step, the AI system searches for and pulls relevant information from an external source. This could be anything from databases, indexed documents, or even live search engines.</p><p>Traditional search engines rely heavily on keyword matching, but RAG improves upon this by considering semantic search. This means the system doesn’t just look for literal matches; it also understands the context and meaning behind the query to return the most relevant information.</p><p><strong>Generation</strong>:<br>After retrieving the most pertinent information, the generative model uses that data to produce natural language responses or content. The generative model, such as a transformer-base.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4d6a371a30d9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Unlocking the Power of LangChain: Building Smarter AI Applications]]></title>
            <link>https://anilvermaspeaks.medium.com/unlocking-the-power-of-langchain-building-smarter-ai-applications-9f79678c65cf?source=rss-41d4d241429------2</link>
            <guid isPermaLink="false">https://medium.com/p/9f79678c65cf</guid>
            <category><![CDATA[claude]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[langchain]]></category>
            <category><![CDATA[langgraph]]></category>
            <category><![CDATA[chatgpt]]></category>
            <dc:creator><![CDATA[Anil Verma]]></dc:creator>
            <pubDate>Sun, 24 Aug 2025 06:25:05 GMT</pubDate>
            <atom:updated>2025-08-24T06:27:22.855Z</atom:updated>
            <content:encoded><![CDATA[<p>Bridging the gap between <strong>raw AI capabilities</strong> and <strong>useful, user-friendly applications</strong></p><p><strong>Artificial Intelligence (AI)</strong> has made huge leaps in recent years, and tools like ChatGPT or Claude show us just how powerful Large Language Models (LLMs) can be. But if you’ve ever tried to use an LLM for a real-world project, you’ll notice a gap:</p><p>➡️ The model is smart, but it doesn’t remember past conversations.<br> ➡️ It doesn’t know your private data.<br> ➡️ It can’t decide when to call an API or look up information.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/0*GYcMHPgXCxkuxEhW.jpg" /></figure><h4>This is where <strong>LangChain</strong> comes in</h4><p><strong>LangChain </strong>is quickly becoming a <strong>standard framework</strong> for LLM applications. As models become more powerful and businesses push for real-world adoption.</p><p>it’s a movement toward smarter, more integrated AI systems. Whether you’re building a chatbot, a data analysis engine, or an AI-powered workflow, LangChain gives you the scaffolding to make your ideas real.</p><p><strong>Use LangChain to:<br>1. </strong>Add <strong>memory</strong> (so your chatbot remembers what you said earlier).<br>2 . Connect to data sources (like PDFs, databases, or websites).<br>3. Give the AI <strong>tools</strong> (like a calculator, search engine, or API).<br>4. Create <strong>workflows</strong> where multiple steps happen automatically.</p><h4>Key Building Blocks of LangChain</h4><p>To make things clearer, let’s break down the <strong>main components</strong>:</p><h4>1. LLMs</h4><p>The base model (like GPT-4, Claude, or LLaMA). LangChain doesn’t replace the model — it orchestrates it.</p><h4>2. Chains</h4><p>A “chain” is simply a sequence of steps. For example:</p><p><strong>User asks a question → LLM queries a database → LLM formats an answer.</strong></p><h4>3. Agents</h4><p>Agents let the model decide <strong>which tool to use and when</strong>. Example: If you ask, <em>“What’s the square root of 81?”</em>, it might use a calculator instead of guessing.</p><h4>4. Memory</h4><p>LLMs normally forget everything after one response. LangChain adds <strong>short-term or long-term memory</strong>, so conversations feel natural.</p><h4>5. Retrieval (RAG)</h4><p>LangChain can connect to <strong>external data</strong> (your files, company documents, websites). This means your chatbot isn’t limited to what the model was trained on.</p><pre>                                                     ┌───────────────────────<br>                           │     User Interface    │<br>                           │  (Chatbot, Web App,   │<br>                           │   CLI, Slack, etc.)   │<br>                           └──────────┬────────────┘<br>                                      │<br>                                      ▼<br>                           ┌───────────────────────┐<br>                           │       LangChain       │<br>                           │   (The Orchestrator)  │<br>                           └──────────┬────────────┘<br>       ┌──────────────────────────────┼──────────────────────────────┐<br>       │                              │                              │<br>       ▼                              ▼                              ▼<br>┌──────────────┐             ┌──────────────────┐           ┌────────────────────┐<br>│   Memory     │             │  Chains &amp; Agents │           │ Retrieval Modules  │<br>│ (short/long- │             │ (decision logic, │           │ (RAG: connect to   │<br>│   term)      │             │ tool selection)  │           │ DBs, PDFs, APIs)   │<br>└───────┬──────┘             └─────────┬────────┘           └─────────┬──────────┘<br>        │                              │                              │<br>        ▼                              ▼                              ▼<br>┌──────────────┐             ┌──────────────────┐           ┌────────────────────┐<br>│   LLMs       │             │ External APIs &amp;  │           │ Vector DBs /       │<br>│ (GPT, Claude,│             │ Tools (Search,   │           │ Knowledge Bases     │<br>│  LLaMA, etc.)│             │ Calculator, etc.)│           │ (Pinecone, FAISS)  │<br>└──────────────┘             └──────────────────┘           └────────────────────┘<br><br>                                      │<br>                                      ▼<br>                           ┌───────────────────────┐<br>                           │     Final Response     │<br>                           │  (Answer, Action, or  │<br>                           │     Generated Text)   │<br>                           └───────────────────────┘<br><br></pre><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9f79678c65cf" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Biases in OpenAI and DeepSeek]]></title>
            <link>https://anilvermaspeaks.medium.com/biases-in-openai-and-deepseek-cadd03c65262?source=rss-41d4d241429------2</link>
            <guid isPermaLink="false">https://medium.com/p/cadd03c65262</guid>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[generative-ai-tools]]></category>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[deepseek]]></category>
            <dc:creator><![CDATA[Anil Verma]]></dc:creator>
            <pubDate>Sun, 23 Feb 2025 06:08:40 GMT</pubDate>
            <atom:updated>2025-02-23T06:08:40.584Z</atom:updated>
            <content:encoded><![CDATA[<p>Both OpenAI and DeepSeek have their strengths and weaknesses, particularly when it comes to biases. OpenAI’s models are more established but face ongoing challenges with bias mitigation. DeepSeek offers a cost-effective alternative but raises concerns about censorship and data handling<strong>.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*NnZQQGe9Wweb3bPg.png" /></figure><h3>OpenAI</h3><p>OpenAI has been a prominent player in the AI field, known for its advanced models like GPT-3 and GPT-4. However, it has faced criticism for biases in its models. These biases often stem from the data used to train the models, which can include societal biases present in the text data. OpenAI has been working on mitigating these biases, but it’s an ongoing challenge. Additionally, OpenAI has faced scrutiny over its handling of user data and transparency in its operations.</p><h3>DeepSeek</h3><p>DeepSeek, a newer entrant from China, has quickly gained attention for its cost-effective and efficient AI models. However, DeepSeek has also faced criticism, particularly regarding censorship and data handling due to its compliance with Chinese government regulations. This raises concerns about the potential biases in its models, especially in terms of content moderation and the information it provides.</p><h3>Key Differences</h3><ol><li><strong>Data Handling and Privacy:</strong> OpenAI operates under US regulations, which have their own set of privacy and data handling standards. DeepSeek, on the other hand, must comply with Chinese regulations, which can lead to different approaches to data privacy and censorship.</li><li><strong>Cost and Accessibility:</strong> DeepSeek offers a more affordable and open-source model compared to OpenAI’s proprietary models, making it accessible to a broader audience.</li><li><strong>Bias Mitigation Efforts:</strong> Both companies are working on reducing biases in their models, but the effectiveness and transparency of these efforts can vary.</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cadd03c65262" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Understanding Data Types: Structured, Unstructured, and Semi-Structured Data]]></title>
            <link>https://anilvermaspeaks.medium.com/understanding-data-types-structured-unstructured-and-semi-structured-data-971d15f3331e?source=rss-41d4d241429------2</link>
            <guid isPermaLink="false">https://medium.com/p/971d15f3331e</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[node]]></category>
            <category><![CDATA[aws]]></category>
            <dc:creator><![CDATA[Anil Verma]]></dc:creator>
            <pubDate>Sat, 22 Feb 2025 11:47:16 GMT</pubDate>
            <atom:updated>2025-02-22T11:47:16.434Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Fhcnm52EVxxP-DBC" /></figure><p>Data is at the core of every digital system, driving decision-making, analytics, and automation. However, not all data is created equal. It exists in different forms, categorized broadly into three types: <strong>Structured Data, Unstructured Data, and Semi-Structured Data</strong>. Understanding these categories is crucial for data engineers, analysts, and developers working with databases, big data, and AI models.</p><h3>1. Structured Data</h3><p>Structured data is organized and stored in a fixed format within databases, making it easily searchable and analyzable. It follows a predefined schema, usually consisting of rows and columns, similar to spreadsheets or relational databases (SQL).</p><h3>Characteristics of Structured Data:</h3><ul><li>Highly organized and stored in tabular format</li><li>Follows a strict schema (e.g., database tables)</li><li>Easily searchable using query languages like SQL</li><li>Typically stored in relational databases (MySQL, PostgreSQL, Oracle, etc.)</li></ul><h3>Examples:</h3><ul><li>Customer information in an e-commerce database (Name, Email, Order History)</li><li>Financial transactions stored in banking systems</li><li>Inventory management data in a retail database</li></ul><h3>2. Unstructured Data</h3><p>Unstructured data lacks a predefined format and is not stored in traditional database structures. It constitutes the majority of data generated today, including multimedia files, social media posts, and IoT sensor outputs.</p><h3>Characteristics of Unstructured Data:</h3><ul><li>Does not have a predefined model or schema</li><li>Difficult to store and manage in relational databases</li><li>Requires specialized tools like NoSQL databases, data lakes, or AI-driven search mechanisms</li><li>Examples include text, images, videos, and audio files</li></ul><h3>Examples:</h3><ul><li>Social media posts (tweets, Facebook updates, Instagram stories)</li><li>Emails and customer feedback surveys</li><li>Audio recordings from customer service calls</li><li>Video content from security cameras or YouTube</li></ul><h3>3. Semi-Structured Data</h3><p>Semi-structured data lies between structured and unstructured data. While it doesn’t follow a strict tabular format, it contains metadata or tags that provide some organization.</p><h3>Characteristics of Semi-Structured Data:</h3><ul><li>Contains some organizational properties but lacks rigid structure</li><li>Often stored in formats like JSON, XML, or YAML</li><li>More flexible than structured data but easier to process than unstructured data</li><li>Used in modern applications for data exchange and APIs</li></ul><h3>Examples:</h3><ul><li>JSON and XML files used in web APIs</li><li>Emails (structured headers + unstructured body text)</li><li>Sensor data with metadata (e.g., temperature readings with timestamps)</li><li>NoSQL databases (MongoDB, Cassandra)</li></ul><h3>Choosing the Right Data Type for Your Needs</h3><p>Understanding the type of data helps in selecting the right storage, retrieval, and processing mechanisms.</p><ul><li>Use <strong>relational databases (SQL)</strong> for structured data where consistency and transactional integrity matter.</li><li>Opt for <strong>NoSQL, cloud storage, or data lakes</strong> for unstructured data like images, videos, and logs.</li><li>Work with <strong>semi-structured formats</strong> like JSON or XML for flexible data exchange in APIs and modern applications.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=971d15f3331e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Unleashing Creativity: How Generative AI is Shaping the Future]]></title>
            <link>https://anilvermaspeaks.medium.com/unleashing-creativity-how-generative-ai-is-shaping-the-future-5f637cdf1b4b?source=rss-41d4d241429------2</link>
            <guid isPermaLink="false">https://medium.com/p/5f637cdf1b4b</guid>
            <category><![CDATA[generative-ai-tools]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[ai-technology]]></category>
            <category><![CDATA[content-creation]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Anil Verma]]></dc:creator>
            <pubDate>Mon, 06 Jan 2025 13:25:56 GMT</pubDate>
            <atom:updated>2025-01-06T13:25:56.432Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4dVZ7nasNU1wxcoZTuwaIA.png" /></figure><p>Generative AI, or generative artificial intelligence, is a type of AI that can create new content like text, images, music, audio, and videos. It uses large AI models, called foundation models, that can perform tasks like summarisation, Q&amp;A, and classification.</p><p>Generative AI, Have the potential to drastically change the way we approach content creation. Gen AI has made remarkable progress, with new tools, regulations, and technological advancements emerging every month. While ChatGPT and AI advancements in general have sparked fear among some, it’s evident that machine learning holds significant potential for positive impact.</p><h4>Machine learning is a branch of artificial intelligence (AI)</h4><p>Machine learning is a branch of artificial intelligence (AI) that focuses on creating systems capable of learning and adapting from data without explicit programming. Instead of following predefined instructions, machine learning models analyze data patterns to make predictions, decisions, or classifications autonomously.</p><p><strong>Explanation:</strong></p><ul><li><strong>Core of AI:</strong> Machine learning serves as a foundational approach to building intelligent systems.</li><li><strong>Data-Driven Learning:</strong> Models are trained on datasets to uncover patterns, trends, and relationships, enabling them to perform tasks such as recognizing images, predicting outcomes, or personalizing experiences.</li><li><strong>What it is:</strong> A way for computers to get smarter by learning from examples.</li><li><strong>How it works:</strong> Think of it like teaching a child — if you show them enough pictures of cats and dogs, they eventually learn to tell the difference without needing step-by-step rules.</li><li><strong>Why it’s special:</strong> The more data the system sees, the better it gets at understanding and solving problems on its own.</li></ul><p>Happy Learning 👏 👏 👏</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5f637cdf1b4b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Deep Dive into Angular Signals: Writable, Readable Signals, Zone.js, and RxJS]]></title>
            <link>https://anilvermaspeaks.medium.com/a-deep-dive-into-angular-signals-writable-readable-signals-zone-js-and-rxjs-6dd2f108eb31?source=rss-41d4d241429------2</link>
            <guid isPermaLink="false">https://medium.com/p/6dd2f108eb31</guid>
            <category><![CDATA[angular]]></category>
            <category><![CDATA[rxjs]]></category>
            <category><![CDATA[node]]></category>
            <category><![CDATA[react]]></category>
            <category><![CDATA[javascript]]></category>
            <dc:creator><![CDATA[Anil Verma]]></dc:creator>
            <pubDate>Wed, 18 Dec 2024 08:35:16 GMT</pubDate>
            <atom:updated>2024-12-18T08:35:16.687Z</atom:updated>
            <content:encoded><![CDATA[<p>New reactive programming paradigm aimed at simplifying state management within Angular applications</p><h4>What Are Angular Signals?</h4><p>At their core, <strong>Angular signals</strong> are a new reactive programming paradigm aimed at simplifying state management within Angular applications. They are <strong>reactive values</strong> that automatically update when their dependencies change, but with a more straightforward and lightweight syntax than using traditional <strong>RxJS observables</strong>. Signals can be either <strong>writable</strong> or <strong>readable</strong>, and they are used to manage state in a more declarative way.</p><ul><li><strong>Writable signals</strong> allow you to <strong>mutate</strong> state, meaning you can both <strong>read</strong> and <strong>write</strong> to the signal.</li><li><strong>Readable signals</strong> are <strong>immutable</strong> — you can only <strong>read</strong> the value, which is usually computed based on other signals or external data sources.</li></ul><h4>Why Signals?</h4><p>While RxJS provides a powerful mechanism for handling state and events in Angular, it comes with its own complexity and learning curve. Angular signals offer a simpler alternative, allowing for more intuitive state management with fewer abstractions. You no longer need to manage subscriptions and operators like map(), filter(), and mergeMap(). Instead, signals give you direct access to state changes, making reactivity more explicit and easier to handle.</p><h4>How Signals Improve Change Detection</h4><p>Unlike Zone.js-based change detection, Angular signals have <strong>fine-grained reactivity</strong>. This means that Angular only reacts to changes in the specific signals you are working with, not all async operations globally. Signals don’t require Zone.js to track every asynchronous event — instead, Angular uses <strong>smart subscriptions</strong> to trigger updates only when relevant signals change.</p><p>In simpler terms, when a writable signal updates, Angular knows exactly where the change occurred, and it re-renders only the components or parts of the template that are <strong>directly dependent on</strong> that signal. This makes it possible to achieve more efficient change detection and avoids unnecessary re-renders that can happen with Zone.js.</p><h4><strong>Creating and Using Writable Signals</strong></h4><p>Writable signals are created using the signal() function, which allows you to define an initial value. You can then use the .set() method to modify the value of the signal, and the signal’s new value will automatically propagate through the app.</p><pre>import { Component } from &#39;@angular/core&#39;;<br>import { signal } from &#39;@angular/core&#39;;<br><br>@Component({<br>  selector: &#39;app-counter&#39;,<br>  template: `<br>    &lt;div&gt;<br>      &lt;p&gt;Counter: {{ counter() }}&lt;/p&gt;<br>      &lt;button (click)=&quot;increment()&quot;&gt;Increment&lt;/button&gt;<br>      &lt;button (click)=&quot;decrement()&quot;&gt;Decrement&lt;/button&gt;<br>    &lt;/div&gt;<br>  `,<br>})<br>export class CounterComponent {<br>  counter = signal(0); // Writable signal<br><br>  increment() {<br>    this.counter.set(this.counter() + 1);<br>  }<br><br>  decrement() {<br>    this.counter.set(this.counter() - 1);<br>  }<br>}</pre><ul><li>We create a writable signal counter with an initial value of 0.</li><li>When the increment or decrement buttons are clicked, we modify the signal using the .set() method.</li><li>The component’s view automatically reflects the updated value without requiring manual subscriptions or change detection triggers.</li></ul><h4>Readable Signals: Immutable State in Angular</h4><p>Readable signals are used to store <strong>immutable</strong> values that <strong>cannot</strong> be changed directly. These signals are <strong>computed</strong> or <strong>derived</strong> from other signals, making them ideal for cases where you want to expose a value that depends on other parts of the state.</p><pre>import { Component } from &#39;@angular/core&#39;;<br>import { signal } from &#39;@angular/core&#39;;<br><br>@Component({<br>  selector: &#39;app-user&#39;,<br>  template: `<br>    &lt;div&gt;<br>      &lt;p&gt;Full Name: {{ fullName() }}&lt;/p&gt;<br>    &lt;/div&gt;<br>  `,<br>})<br>export class UserComponent {<br>  firstName = signal(&#39;John&#39;);<br>  lastName = signal(&#39;Doe&#39;);<br><br>  fullName = signal(() =&gt; `${this.firstName()} ${this.lastName()}`); // Readable signal<br>}</pre><ul><li>fullName is a <strong>readable</strong> signal that computes its value based on the firstName and lastName signals.</li><li>The fullName signal can’t be changed directly — it’s always derived from the other two signals.</li></ul><h4>Writable Signals</h4><ul><li><strong>Mutable</strong> state — can be modified using the .set() or .update()method.</li><li>Ideal for <strong>local component state</strong> or <strong>global state</strong> that needs to change over time.</li></ul><h4>Readable Signals</h4><ul><li><strong>Immutable</strong> state — cannot be modified directly.</li><li>Ideal for <strong>computed values</strong> or <strong>derived state</strong> that depends on other signals.</li></ul><h4>Zone.js vs Signals</h4><ul><li><strong>Zone.js</strong> is best suited for scenarios where you need to handle <strong>global asynchronous events</strong> (like HTTP requests, timers, etc.) and automatically trigger change detection across the entire app.</li><li><strong>Signals</strong> offer <strong>fine-grained reactivity</strong>, allowing Angular to react only to specific changes. Signals are much more efficient and predictable for <strong>local state management</strong> within a component or service.</li></ul><h4><strong>RxJS vs Signals</strong></h4><ul><li><strong>RxJS</strong> is still incredibly powerful for handling complex asynchronous streams, event handling, and multi-step data flows.</li><li><strong>Signals</strong>, on the other hand, offer a <strong>simpler, more declarative</strong> approach for managing state that needs to update reactively. Signals don’t replace RxJS; they complement it for <strong>local state</strong> and <strong>derived state</strong>.</li></ul><p>Happy Learning……👏👏👏</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6dd2f108eb31" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AWS Accounts(Multiple AWS Accounts VS Single-Account)]]></title>
            <link>https://anilvermaspeaks.medium.com/aws-accounts-multiple-aws-accounts-vs-single-account-20d73cba3709?source=rss-41d4d241429------2</link>
            <guid isPermaLink="false">https://medium.com/p/20d73cba3709</guid>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[aws-certification]]></category>
            <category><![CDATA[s3-bucket]]></category>
            <category><![CDATA[architecture]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <dc:creator><![CDATA[Anil Verma]]></dc:creator>
            <pubDate>Sat, 05 Oct 2024 03:28:01 GMT</pubDate>
            <atom:updated>2024-10-05T03:28:01.122Z</atom:updated>
            <content:encoded><![CDATA[<p>In AWS, an account serves as a container for your cloud resources and services. Each AWS account is distinct, with its own security settings, resources, and billing.</p><p>When you sign up for AWS, you create an account that allows you to access and manage various AWS services. Each account is uniquely identified and operates independently from other accounts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/661/0*dt1DfZqvSuS2IxGr.png" /></figure><h3>Key Features of AWS Accounts</h3><ol><li><strong>Isolation</strong>: Each account is isolated from others, meaning resources, permissions, and security settings do not overlap. This isolation enhances security, making it easier to manage and control access.</li><li><strong>Billing and Cost Management</strong>: Each account has its own billing structure, allowing you to track costs and usage separately. This is particularly useful for organizations with different departments or projects, enabling precise cost allocation.</li><li><strong>Resource Management</strong>: Resources such as EC2 instances, S3 buckets, and RDS databases are tied to a specific account. This organization helps maintain clarity in resource allocation and management.</li><li><strong>Security Policies</strong>: AWS accounts use Identity and Access Management (IAM) to define who can access resources and what actions they can perform. You can create users, roles, and policies specific to each account.</li><li><strong>Compliance</strong>: Having multiple accounts can help organizations meet different regulatory requirements. Each account can be configured to adhere to specific compliance standards.</li></ol><h3>Single-Account Setup VS Multiple AWS Accounts</h3><h4>single AWS account</h4><p>While having a single AWS account might seem simpler, it comes with its own set of challenges:</p><p><strong>1. Increased Security Risks</strong><br>A single account means that all resources are under one roof. If that account is compromised, it puts all resources at risk, making it a bigger target for attackers.</p><p><strong>2. Limited Resource Management</strong><br>Managing resources and permissions can become complicated. As more teams and projects are added, it can be hard to keep track of who has access to what.</p><p><strong>3. Compliance Challenges</strong><br>Different projects may have unique compliance needs. In a single-account setup, it’s difficult to apply tailored policies and controls for each project, increasing the risk of non-compliance.</p><h4>Multiple AWS Accounts</h4><p>Imagine a data analytics team in Account A that needs to access a database in Account B. By setting up a cross-account IAM role, the team can easily access the database without needing separate accounts in Account B.</p><p><strong>1. Improved Security</strong><br>Separating workloads into different accounts reduces the risk of security breaches. If one account is compromised, the others remain protected.</p><p><strong>2. Better Resource Management</strong><br>Having separate accounts allows teams to manage their resources independently, leading to more efficient use of resources tailored to specific needs.</p><p><strong>3. Enhanced Compliance</strong><br>Different teams or projects may have varying compliance requirements. Using multiple accounts helps in managing compliance more effectively, as each account can have tailored policies and controls.</p><p><strong>4. Simplified Billing</strong><br>Multi-account setups can help with better billing and cost management. You can easily track costs per account, making it simpler to understand where money is being spent.</p><p><strong>5. Isolation of Environments<br></strong>Using different accounts for development, testing, and production helps in isolating environments. This minimizes risks, as changes in one environment won’t impact others.</p><h4>How It Works:</h4><ol><li><strong>Account A</strong>: The account where the user or service is located.</li><li><strong>Account B</strong>: The account that has the resources you want to access.</li><li><strong>IAM Role</strong>: This is set up in Account B and includes:</li><li><strong>Trust Policy</strong>: This tells who from Account A can use the role.</li><li><strong>Permissions Policy</strong>: This defines what can be accessed.</li></ol><p><strong>Step-by-Step Process</strong></p><ol><li><strong>Request Access</strong>: A user in Account A asks to use the role in Account B.</li><li><strong>AssumeRole API Call</strong>: The user makes a request to AWS to assume the role.</li><li><strong>Get Temporary Access</strong>: AWS gives temporary access credentials.</li><li><strong>Access Resources</strong>: The user can now use these credentials to access resources in Account B.</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=20d73cba3709" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Real User Monitoring(RUM)]]></title>
            <link>https://anilvermaspeaks.medium.com/real-user-monitoring-rum-aa96d5858788?source=rss-41d4d241429------2</link>
            <guid isPermaLink="false">https://medium.com/p/aa96d5858788</guid>
            <category><![CDATA[react]]></category>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[debugging]]></category>
            <category><![CDATA[performance]]></category>
            <category><![CDATA[automation]]></category>
            <dc:creator><![CDATA[Anil Verma]]></dc:creator>
            <pubDate>Sun, 12 May 2024 06:11:59 GMT</pubDate>
            <atom:updated>2024-05-12T06:11:59.732Z</atom:updated>
            <content:encoded><![CDATA[<p>Tracks and measures the end user experience in an application, including App performance, and how long API calls taking time.<br>it provides actionable, data-driven details to improve website’s end user experience.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/0*SZh7ZbPEMJRzjuTj" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/710/0*dwa7kujSTXEkoG0y" /></figure><h4><strong>APM (Application Performance Monitoring/ Real user monitoring tool)—</strong></h4><p>it helps in getting insights of web application performance with the help of multiple monitoring capabilities</p><ol><li><strong>Real Time Visibility — </strong>quick summary of how your website is performing in real-time by keeping track of key metrics such as Response Time, Page load time, Page Views, throughput</li><li><strong>User Experience Across Geographies — </strong>Visualize how website is performing across countries on world map. it helps in quickly identifying locations where users are affected with website performance by analysing response time, error count, and throughput.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/0*5K4aE1hbXIDSnkbT" /></figure><ol><li><strong>Detect sluggish transactions in no-time -</strong>have a close eyes on the number of page views for website as well as the errors obtained for individual transactions to identify any potential issues.</li><li><strong>Track user sessions- </strong>Get<strong> </strong>the number of user sessions that have connected to website. Also get status of each session to determine the duration and quality of user sessions established.</li><li><strong>Detect script errors-</strong>Identify JS errors that are impacting website’s performance in real time.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/0*8BNNM8Kx1me0KJij" /></figure><p><strong>Monitoring solutions such as New Relic come with built-in dashboards.</strong> We can also build our own custom dashboards based on the metrics that matters most to us. <br>other tools — AppDynamics Browser RUM, Retrace, Datadog</p><h4><strong>How real user monitoring works</strong></h4><p>Real user monitoring works by injecting code into an application to capture metrics while the application is in use. Client side applications are monitored by injecting Javascript code.</p><p>Native mobile applications are monitored by adding the monitoring library directly to your mobile application package.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aa96d5858788" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A/B Testing or Split Testing]]></title>
            <link>https://anilvermaspeaks.medium.com/a-b-testing-or-split-testing-bc3da0d8b3a3?source=rss-41d4d241429------2</link>
            <guid isPermaLink="false">https://medium.com/p/bc3da0d8b3a3</guid>
            <category><![CDATA[testing]]></category>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[nodejs]]></category>
            <category><![CDATA[a-b-testing]]></category>
            <category><![CDATA[architecture]]></category>
            <dc:creator><![CDATA[Anil Verma]]></dc:creator>
            <pubDate>Sat, 11 May 2024 04:20:19 GMT</pubDate>
            <atom:updated>2024-05-11T04:20:19.578Z</atom:updated>
            <content:encoded><![CDATA[<p>Compare two versions of an webpage to identify the better performer. based on real data.</p><p><strong>In A/B testing</strong><br> <strong>A —</strong> refers to ‘control’ or the original part of webpage. <br><strong>B —</strong> refers to ‘variation’ or a new version.</p><p><strong>Idea</strong>- A/B testing is one of main component into process of Conversion Rate Optimization (CRO), using which we can gather user insights. we can use this also for collecting data to understand user behavior, engagement rate, pain points.</p><h3>Different types of A/B tests</h3><p><strong>Split URL testing — </strong>When you run a Split URL test, your website traffic is split between two url’s first one as original url, second url with variations.<br> this is recommended testing method for dynamic content.</p><p><strong>Multipage testing — </strong>Test changes to particular elements across multiple pages. based on this you can analyse impact across an entire flow. It enables to implement the same change across pages to ensure that web page visitors get consistent experiences.</p><p><strong>Multivariate testing- </strong>variations of multiple page variables are simultaneously tested to analyze which combination is performing better. it helps in avoiding the need to conduct several sequential A/B tests and saves time.</p><h3>Approaches to run an A/B test-</h3><p><strong>Frequentist Approach — </strong>Requires test to run for a set period to get correct data from it but can’t figure out how close or far A and B actually are.<br>data used from current experiment to draw conclusions.</p><p><strong>Bayesian Approach- </strong>You can now plan better, have a more accurate reason to end tests. use existing data to draw conclusions. the Bayesian approach provides actionable results almost 50% faster then <strong>Frequentist Approach</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bc3da0d8b3a3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Git Branch Strategies]]></title>
            <link>https://anilvermaspeaks.medium.com/git-branch-strategies-695da0437820?source=rss-41d4d241429------2</link>
            <guid isPermaLink="false">https://medium.com/p/695da0437820</guid>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[nodejs]]></category>
            <category><![CDATA[github]]></category>
            <category><![CDATA[git]]></category>
            <category><![CDATA[react]]></category>
            <dc:creator><![CDATA[Anil Verma]]></dc:creator>
            <pubDate>Fri, 29 Mar 2024 09:11:19 GMT</pubDate>
            <atom:updated>2024-03-29T09:11:19.021Z</atom:updated>
            <content:encoded><![CDATA[<p>Organize code and reduce the time needed to manage versioning</p><p>there are four most popular branching strategies — <br>1. Git-Flow<br>2. GitHub-Flow<br>3. Trunk</p><h4>Git-Flow-</h4><p>The main idea behind the git-flow is to isolate your work into different types of branches. There are few famous branch types as below. life-cycle involves branching off from develop, integrating features, creating release branches for testing, merging into main/master</p><ul><li>Main</li><li>Develop</li><li>Feature</li><li>Release</li><li>Hotfix</li></ul><p><strong>Pros:</strong><br>- Good for large teams and aligning work across multiple teams.<br><strong>Cons:</strong><br>- with complexity in project, git flow slow the development process and release cycle.</p><h4>GitHub-Flow:</h4><p>it simplifies Git-Flow by eliminating release branches. It revolves around one active development branch (main or master) that is directly deployed to production.<br><strong>Pros:</strong><br>- Faster feedback cycles and shorter production cycles.<br>- Ideal for asynchronous work in smaller teams.</p><p><strong>Cons:</strong><br>- Long-living branches can complicate the process.</p><h4>Trunk Based Development</h4><p>Trunk Based Development promotes a single shared branch called “trunk” and eliminates long-living branches.<br>smaller teams commit directly to the trunk, while larger teams create short-lived feature branches.</p><p><strong>Pros:</strong><br>- Enhances collaboration and reduces merge conflicts.</p><p><strong>Cons:</strong><br>- Requires an experienced team that can slice features appropriately for regular integration.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=695da0437820" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>