<?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 prachi lal on Medium]]></title>
        <description><![CDATA[Stories by prachi lal on Medium]]></description>
        <link>https://medium.com/@plalindia01?source=rss-2a468d39b08f------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*9GgRFpvPVUlapLPKYm4czw.jpeg</url>
            <title>Stories by prachi lal on Medium</title>
            <link>https://medium.com/@plalindia01?source=rss-2a468d39b08f------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 01:27:20 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@plalindia01/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[Who am I?]]></title>
            <link>https://medium.com/@plalindia01/who-am-i-126fd8617841?source=rss-2a468d39b08f------2</link>
            <guid isPermaLink="false">https://medium.com/p/126fd8617841</guid>
            <category><![CDATA[introducing-myself]]></category>
            <category><![CDATA[introduction]]></category>
            <dc:creator><![CDATA[prachi lal]]></dc:creator>
            <pubDate>Thu, 29 May 2025 11:00:57 GMT</pubDate>
            <atom:updated>2025-05-29T13:33:31.459Z</atom:updated>
            <content:encoded><![CDATA[<p>Can’t believe I’m actually writing a “Who am I?” post on Medium. I remember <em>dreading</em> publishing my first write-up… but here we are.</p><p>The facts:</p><ul><li>Born in 2001</li><li>Pronouns: She/Her</li><li>Bachelor of Technology graduate from Symbiosis International University (2024)</li></ul><h4>What am I doing here?</h4><p>To be honest, I have no clue.</p><p>Am I here for fame? Bruh, please… I got like 10 people reading my stuff (whom I really really love and appreciate).</p><p>Am I here because <em>I’m the wielder of the Elder Wand</em>? Maybe.</p><p>As I build random stuff- bots, APIs, dashboards, dumb side projects that accidentally become cool, this weird internal narrator pops up and gives some pretty spicy takes. Normally, I’d dump them in a journal. But I guess I overcame my shyness with <em>just a little</em> validation and also, it’s 2025.</p><h4>What to expect from this page?</h4><p>Tech rants (sometimes funny, sometimes depressed)</p><p>Quick builds I’m proud of</p><p>Unfiltered thoughts. I’m a huge fan of <strong>unstructured, unmoderated debate</strong>.</p><p>Also, I can handle my drinks. So if you drop a hot take, I guarantee some witty replies and sharp comebacks. XD</p><p>I can handle git, in case you down to fork.</p><p>I play <strong>Valorant</strong>, and working at <strong>Riot Games</strong> is a dream.</p><p>I’d love to go on about how I’m in love with the ocean, how I self-validate and overcompensate for late working hours by buying myself overpriced coffee…<br>But hey, let’s leave <em>some</em> mystery between us.</p><h4>Credentials (in case you want to stalk me):</h4><ul><li><strong>LinkedIn: </strong><a href="https://www.linkedin.com/in/prachi-lal/">Prachi Lal</a></li><li><strong>Portfolio:</strong> <a href="https://fish-turn-db4.notion.site/Prachi-Lal-APIs-Automation-and-Clean-Data-Handled-1f4ebdfd0328807994ebecf4dbed27a0"><em>Prachi Lal: APIs, Automation, and Clean Data — Handled</em></a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=126fd8617841" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why I Let My Database Think for Me (And You Should Too)]]></title>
            <link>https://medium.com/@plalindia01/why-i-let-my-database-think-for-me-and-you-should-too-0520326b3814?source=rss-2a468d39b08f------2</link>
            <guid isPermaLink="false">https://medium.com/p/0520326b3814</guid>
            <category><![CDATA[database-design]]></category>
            <category><![CDATA[database]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[twitter]]></category>
            <dc:creator><![CDATA[prachi lal]]></dc:creator>
            <pubDate>Mon, 19 May 2025 04:14:54 GMT</pubDate>
            <atom:updated>2025-05-19T04:14:54.036Z</atom:updated>
            <content:encoded><![CDATA[<p>Let’s face it — we’ve all been there. Scoffing at the idea of embedding logic in the database.</p><blockquote>Business logic belongs in the application layer</blockquote><p>they say, sipping their artisanal coffee, pushing their microservices agenda like it’s gospel.</p><p>But what if …just what if- our database systems, armed with procedural language capabilities, are more than just glorified data vaults? What if they’re the unsung heroes, the real workhorses, while our applications prance around in React and Tailwind, looking pretty?</p><p>I say- let’s test when it breaks. Let’s trust my database in shining armor. The application? Just an ornament. A wrapper. A UX sugar cube. The database? That’s where the heavy lifting happens.</p><p>The Benefit?<br>Every time I ship a new version of my app, I don’t have to worry about my “global” logic. It’s embedded in the database, safe and sound, versioned and battle-tested. My app can change its clothes every week, but the soul — the logic — remains untouched.</p><p>The Cost?<br>Database migrations might require someone who can handle their drink. Preferably neat. With a twist of caution and a splash of rollback strategy.</p><p>Still not convinced?</p><p>Let me build something for y’all. Something universal. Something that transcends frameworks, languages, and even your favorite VS Code theme.</p><p>Drumroll, please…</p><p><strong><em>An Identity String.</em></strong></p><p>Yes, I hear you already:</p><blockquote>Prachi, just use an auto-incremented dtype or something and spare us the time.</blockquote><p>With a sly smile, I’d say-</p><blockquote>Sit down, friend. And listen to what <a href="https://tomoharutsutsumi.medium.com/why-auto-increment-ids-are-challenging-for-scalability-bf51533e9632">Tomoharu</a> has to say.</blockquote><p>Look I trust the guy. So I am gonna do as he says. I am gonna build a snowflake id generator and since its 2025 already, I will be using PostgreSQL.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*-HzTWQD67bXcUMNP.png" /></figure><p>Let me be honest and come clean- the concept of snowflake ids was birthed by Twitter and the the code I am producing here, my friend copilot helped me prepare, as coding is a vibe these days.</p><pre>CREATE TABLE IF NOT EXISTS snowflake_state (<br>    last_timestamp BIGINT NOT NULL,<br>    sequence INT NOT NULL<br>);<br><br>INSERT INTO snowflake_state VALUES (0, 0) ON CONFLICT DO NOTHING;</pre><p>This is your database’s diary. It remembers the last time it handed out an ID and how many it gave out in that exact millisecond. Without this, your IDs would be about as unique as usernames on a new social media app.</p><pre>CREATE OR REPLACE FUNCTION generate_snowflake_id(machine_id INT DEFAULT 1)<br>RETURNS BIGINT AS $$<br>DECLARE<br>    epoch BIGINT := 1700000000000; -- Custom epoch (e.g., Nov 2023)<br>    timestamp_ms BIGINT;<br>    last_ts BIGINT;<br>    seq INT;<br>    max_seq INT := 4095; -- 12 bits<br>    result BIGINT;<br>BEGIN<br>    LOOP<br>        timestamp_ms := FLOOR(EXTRACT(EPOCH FROM clock_timestamp()) * 1000);<br><br>        SELECT last_timestamp, sequence INTO last_ts, seq FROM snowflake_state FOR UPDATE;<br><br>        IF timestamp_ms = last_ts THEN<br>            seq := seq + 1;<br>            IF seq &gt; max_seq THEN<br>                -- Wait for next millisecond<br>                PERFORM pg_sleep(0.001);<br>                CONTINUE;<br>            END IF;<br>        ELSE<br>            seq := 0;<br>        END IF;<br><br>        UPDATE snowflake_state SET last_timestamp = timestamp_ms, sequence = seq;<br><br>        result := ((timestamp_ms - epoch) &lt;&lt; 22) -- 41 bits<br>                  | ((machine_id &amp; 1023) &lt;&lt; 12)   -- 10 bits<br>                  | (seq &amp; 4095);                 -- 12 bits<br><br>        RETURN result;<br>    END LOOP;<br>END;<br>$$ LANGUAGE plpgsql;<br><br><br>SELECT generate_snowflake_id(42); -- Replace 42 with your machine/node ID</pre><p>This where the magic happens, the main function. Now lets put this in a table as a primary key for something</p><pre>DROP TABLE IF EXISTS users CASCADE;<br>CREATE TABLE users (<br>    id BIGINT PRIMARY KEY DEFAULT generate_snowflake_id(1),<br>    username TEXT NOT NULL,<br>    email TEXT UNIQUE NOT NULL,<br>    created_at TIMESTAMP DEFAULT now()<br>);<br><br>INSERT INTO users (username, email)<br>VALUES (&#39;prachi_dev&#39;, &#39;prachi@example.com&#39;);<br><br>SELECT * FROM users;</pre><p>Now, when I implemented this on my system (yes I do multiple tests before sharing it with y’all) this is the output I got:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/941/1*jDkIbx-2dE_1nz6C0uDlFQ.png" /></figure><p>There you have it, your very own Snowflake id generator, handcrafted in PostgreSQL. Next time someone comes to you preaching how logic only belongs to the application layer, sip your <a href="https://www.quora.com/What-is-kadak-chai"><em>kadak chai</em></a> and smile.</p><p>If you’re still here, bless your patience.</p><p>If you liked this article, let me know!<br>If you found this abysmal, let me know!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0520326b3814" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How MongoDB Saved Me from Writing a Thousand Joins]]></title>
            <link>https://medium.com/@plalindia01/how-mongodb-saved-me-from-writing-a-thousand-joins-7d48694d7f99?source=rss-2a468d39b08f------2</link>
            <guid isPermaLink="false">https://medium.com/p/7d48694d7f99</guid>
            <category><![CDATA[sql-vs-nosql]]></category>
            <category><![CDATA[nosql]]></category>
            <category><![CDATA[sql]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[database]]></category>
            <dc:creator><![CDATA[prachi lal]]></dc:creator>
            <pubDate>Tue, 08 Apr 2025 20:15:30 GMT</pubDate>
            <atom:updated>2025-04-08T20:15:30.082Z</atom:updated>
            <content:encoded><![CDATA[<blockquote>I just wanted to store a few tours. Next thing I know, I’m knee-deep in normalization, foreign keys, and existential dread!</blockquote><p>So here’s the deal. I set out to build a simple database system for a <strong>tour booking company</strong>. You know — tours, cities, countries, keywords like “adventurous jungle safari” or “romantic European getaway.” Easy, right?</p><p>Well, buckle up. Because depending on whether you’re in <strong>Camp SQL (PostgreSQL)</strong> or <strong>Camp NoSQL (MongoDB)</strong>, your life is either about to become a peaceful vacation or an unpaid internship in relational database theory.</p><h4>The Idea</h4><p>Here’s what I wanted-</p><p>Each tour should have:</p><ol><li>A list of cities (because apparently people visit more than one place per trip).</li><li>A list of countries (because borders are just lines on a map).</li><li>A bunch of keywords (which could be actual words or full-blown marketing phrases).</li><li>Ability to filter tours easily. Like, give me all tours with the keyword “mountain” or “beach” and that go to Italy.</li></ol><p><strong>A reasonable ask, right?</strong></p><h4><em>Enter MongoDB!</em></h4><p>The database equivalent of yoga pants — <strong>flexible, forgiving, and always ready for brunch</strong>.</p><p>Using MongoDB, I just made a document like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8nxzgVsBBRAIW7NNGks2mw.png" /></figure><h4>PostgreSQL</h4><p>The database for all the control freaks out there. Its like a man/woman screaming with a whip- “You Shall Not Pass Without a Schema”.<br>I started out innocent enough:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qmQ4Jyp85dk01HgsDniKSg.png" /></figure><p>And for a glorious five minutes, I thought, “Hey, this isn’t so bad.” Then came the problems:</p><p><strong><em>❌ Limited Querying</em></strong></p><p>Sure, PostgreSQL supports arrays, but querying them feels like assembling IKEA furniture with a spoon.</p><p>Readable? Barely. Maintainable? Hah. Performant at scale? Only if you sacrifice a goat under a full moon.</p><p><strong><em>❌ Data Integrity</em></strong></p><p>Oh, you want to ensure the country actually <em>exists</em> in a valid list of countries? Good luck doing that with an array column.</p><p><strong><em>❌ Indexing Woes</em></strong></p><p>You <em>can</em> index arrays. But not easily. And not well. And not without watching a 30-minute YouTube video on GIN indexes.</p><h4>The Inevitable Normalization Spiral</h4><p>In the name of data integrity and “best practices,” you soon end up with something like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZQWRlBOw374X60mhyoua7g.png" /></figure><p>Now every time you want to run a simple query, you’re doing 4 joins and questioning your life choices.</p><p>If you’re still here, bless your patience. You were probably hoping for wisdom, but all I’ve got is this emotional damage and a newfound respect for JSON.</p><p>If you liked this article, let me know!<br>If you found this abysmal, let me know!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7d48694d7f99" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Step-by-Step Guide to Hosting a Static Website on AWS]]></title>
            <link>https://medium.com/@plalindia01/step-by-step-guide-to-hosting-a-static-website-on-aws-61d1b705b1fd?source=rss-2a468d39b08f------2</link>
            <guid isPermaLink="false">https://medium.com/p/61d1b705b1fd</guid>
            <category><![CDATA[website-hosting]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[aws-s3]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[prachi lal]]></dc:creator>
            <pubDate>Thu, 03 Apr 2025 10:05:20 GMT</pubDate>
            <atom:updated>2025-04-04T10:09:43.334Z</atom:updated>
            <content:encoded><![CDATA[<p>So, you want to host a static website on AWS? Maybe you’re tired of those “free” website builders that slap their logo everywhere, or maybe you just want to flex your tech skills. Either way, congratulations — you’re about to enter the world of AWS, where simplicity is a myth and documentation feels like a treasure hunt. But don’t worry, I got you. Chug that shot of espresso and let’s get that website live without losing your sanity.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*ln0T9GYxkulYzeAASgm6Hg.gif" /></figure><h3>Step 1: Get Yourself an AWS Account</h3><p>First things first, sign up for an AWS account. If you already have one, great — you’re already past the most frustrating part (billing surprises come later). Just log in to the AWS Management Console. I used the free tier for the purpose of this demonstration, so can you!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/836/1*jcILJnVvnK73oKiyuPYAaw.png" /></figure><h3>Step 2: Create an S3 Bucket</h3><p>You’ll need a fancy file cabinet to hold all your stuff. Creating a bucket is pretty simple stuff, most probably on the first glance, you’ll find the “create bucket” button. Click on that suckah and leave all the options to default except for these 2:<br>1. Object ownership settings to ACLs enabled</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/838/1*RkPjz3YCNNVx28jotmBSuA.png" /></figure><p>2. Uncheck ‘Block Public Access settings for this bucket’ option</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/821/1*IS0_3mRIHsjGf_IsmDJKPA.png" /></figure><p>Let me make your life easier by already letting you know that your bucket name needs to be <strong><em>globally unique</em></strong>. Its like setting up your in game name in Valorant or your Facebook profile. <br>In case your bucket name is not as unique as your personality, you’ll get an error as such.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/827/1*cZxvhWYrOS9vtqY0bDSE5A.png" /></figure><p>AWS will let you know that the name is invalid, but sadly will not give you options.</p><h3>Step 3: Set up your Docs</h3><p>This is the dirty part of the job. This is where you craft the HTML doc for your website. I am not a HTML person, so its ok if you don&#39;t get inspired with what I’ve shared here, but this can be one very boring way to make it work anyway.</p><pre>&lt;!DOCTYPE html&gt;<br>&lt;html&gt;<br>&lt;body&gt;<br><br>&lt;h1&gt;Static Website&lt;/h1&gt;<br><br>&lt;p&gt;Congratulations!  You have configured static website hosting on Amazon S3.&lt;/p&gt;<br><br>&lt;/body&gt;<br>&lt;/html&gt;</pre><blockquote>Break a leg, get your hands dirty.</blockquote><p>Once you are content, with how everything looks, save it as a “.html” file and go ahead to upload in in your S3 bucket.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/826/1*461e_e1gFdkkg6M0segFpA.png" /></figure><h3>Step 4: Some more settings jigsaw</h3><p>You’ll have to edit the ACL and grant read permissions to everyone. Oh yes, the world has been desperately waiting to feast its eyes on your masterpiece.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/832/1*ctR7eZXZe_dZuk77in4pig.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/815/1*K31vxsigMOUE_9XK9koPvw.png" /></figure><h3>Step 5: Declaring what it is</h3><p>Crawl back to the bucket properties and enable the static website hosting options and provide the name for your home page. This is basically telling your cabinet that, “Hey, since you look so good, I am gonna showcase you like a frame”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/828/1*G1PO53e6Q31Cd0Vd3QB-pA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/831/1*_7ENu7XPqu-rX-HFciKkjA.png" /></figure><blockquote>Green flags are the way to go, coding or irl.</blockquote><p>Once you see a successful message like this, you are good to go.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/831/1*NjmWwxSVjUEymafCLmLchg.png" /></figure><h3>Step 6: Test, Celebrate, and Brag</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*b6mP5zyZslmzMC1Aub9fwA.gif" /></figure><p>Your website should now be live! Open a browser and check if everything works. If it doesn’t, welcome to the AWS troubleshooting maze. Otherwise, congrats! You now have a fully functional static website hosted on AWS. Now go flex on your LinkedIn, tweet about it, or just bask in your own tech brilliance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/838/1*mLo7cGdeU-8K4hYSnnb7SQ.png" /></figure><h3>Step 7: Talk to me</h3><p>If you liked this tutorial, let me know!<br>If you found this abysmal, let me know!<br>If you think you can do this better, let me know!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=61d1b705b1fd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Online Flower Shop: Backend Development]]></title>
            <link>https://medium.com/@plalindia01/online-flower-shop-backend-development-7c7f665d5be1?source=rss-2a468d39b08f------2</link>
            <guid isPermaLink="false">https://medium.com/p/7c7f665d5be1</guid>
            <category><![CDATA[sql]]></category>
            <category><![CDATA[backend-development]]></category>
            <category><![CDATA[database-design]]></category>
            <category><![CDATA[postgresql]]></category>
            <category><![CDATA[fastapi]]></category>
            <dc:creator><![CDATA[prachi lal]]></dc:creator>
            <pubDate>Wed, 26 Mar 2025 10:26:59 GMT</pubDate>
            <atom:updated>2025-03-26T10:26:59.027Z</atom:updated>
            <content:encoded><![CDATA[<h4>FastAPI and PostgreSQL</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/917/1*kMjNppvGg6KQnK6S-6Uwzg.png" /></figure><h3>Project Overview</h3><p>The Online Flower Shop API is a web application built using FastAPI and PostgreSQL. It serves as a virtual showcase for different types of flowers, allowing users to register, browse flowers, and place orders online. The application aims to automate the activities related to a florist shop, making online shopping easy and efficient.</p><h3>Features</h3><ul><li>User registration and authentication</li><li>Browse flowers with details</li><li>Place orders for flowers</li><li>Manage stock quantities automatically</li><li>Simple and intuitive API design</li></ul><h3>Technologies Used</h3><ul><li>FastAPI: A modern web framework for building APIs with Python 3.6+ based on standard Python type hints.</li><li>PostgreSQL: A powerful, open-source relational database system.</li><li>Psycopg2: PostgreSQL adapter for Python.</li><li>Uvicorn: ASGI server for running FastAPI applications.</li></ul><h3>Installation</h3><p>Clone the repository:</p><pre>git clone https://github.com/Prachi-Lal/postgres-fastapi-integration.git</pre><p>Create a virtual environment (Optional but recommended):</p><pre>python -m venv venv<br>venv\Scripts\activate</pre><p>Install Required Packages:</p><pre>pip install -r requirements.txt</pre><h3>Database Setup</h3><ol><li>Install PostgreSQL: Make sure you have PostgreSQL installed on your machine. You can download it from PostgreSQL’s official website.</li><li>Run the SQL Scripts: Use the provided SQL scripts to create the necessary tables and insert sample data. You can execute the SQL commands in a PostgreSQL client like psql or any GUI tool like pgAdmin.</li></ol><h3>Running the Application</h3><p>To run the FastAPI application, use the following command:</p><pre>uvicorn main:app --reload</pre><h3>API Endpoints</h3><h3>User Registration</h3><ul><li>Endpoint: /register/</li><li>Method: POST</li><li>Request Body:</li></ul><pre>{<br> “username”: “new_user”,<br> “password”: “password123”,<br> “email”: “new_user@example.com”,<br> “role”: “Customer”<br>}</pre><p>This is how you can use the very useful Swagger UI to test the endpoint.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LkvQvcEO_crxXxoW7LK03Q.png" /></figure><p>A successful run would look like this-</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lrTWHtkd4iCIvBZ-g8sbYQ.png" /></figure><p>Whereas, an unlucky miss would look like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1EcTXPh8dw4bqp7OOdHZTg.png" /></figure><p>For the entire code visit:<br><a href="https://github.com/Prachi-Lal/postgres-fastapi-integration">Here</a></p><p>For a detailed Oracle SQL implementation of an online flower shop, visit:<br><a href="https://github.com/Prachi-Lal/online_flower_shop">Here</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7c7f665d5be1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Battle of wits: Oracle vs. PostgreSQL]]></title>
            <link>https://medium.com/@plalindia01/battle-of-wits-oracle-vs-postgresql-5086499c06da?source=rss-2a468d39b08f------2</link>
            <guid isPermaLink="false">https://medium.com/p/5086499c06da</guid>
            <category><![CDATA[oracle]]></category>
            <category><![CDATA[dbms]]></category>
            <category><![CDATA[postgresql]]></category>
            <category><![CDATA[sql]]></category>
            <category><![CDATA[rdbms]]></category>
            <dc:creator><![CDATA[prachi lal]]></dc:creator>
            <pubDate>Wed, 12 Mar 2025 10:44:20 GMT</pubDate>
            <atom:updated>2025-03-12T14:56:37.882Z</atom:updated>
            <content:encoded><![CDATA[<p>As a part of my recent training in the RDBMS domain, I was taught about 3 superheroes in the arena- Oracle, PostgreSQL and MySQL. This sparked a question in my mind: <em>Which vendor should be used when? </em>While many experts have already explored this, I’ve taken it upon myself to delve deeper and form my own understanding.</p><p>In this piece, I aim to take a shallow dive into some key differences and nuances between Oracle and PostgreSQL.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Cy_7n74TUWZntG5uJB81tw.png" /></figure><p>Before we dive in, let me introduce a simple scoring system: each advantage earns 1 point, while disadvantages are ignored.</p><h3>Let the games begin!</h3><p><strong>Oracle: 0</strong><br><strong>PostgreSQL: 0</strong></p><p>First up — licensing costs. Oracle demands a <em>small fortune</em> (because why not?), while PostgreSQL? <em>Absolutely free.</em> 💸<br> <strong>1 POINT TO POSTGRESQL!</strong></p><p>Next, vendor lock-in. Oracle wants you to stay forever (clingy much?), whereas PostgreSQL lets you walk away anytime, no strings attached.<br> <strong>1 POINT TO POSTGRESQL!</strong></p><p>But wait — Oracle’s optimizer can rewrite queries using a materialized view. PostgreSQL? Nah, it lets you do all the heavy lifting yourself.<br> <strong>1 POINT TO ORACLE!</strong></p><p>Scalar subquery caching? Oracle does it. PostgreSQL doesn’t even try.<br> <strong>1 POINT TO ORACLE!</strong></p><p>Now, let’s talk backup and recovery. Oracle has <strong>RMAN</strong>, incremental, block-level backups, <strong>REDO logs</strong>, and even Point-In-Time Recovery (PITR) like some kind of database time machine.<br> PostgreSQL? <em>pg-dump, pg_basebackup, and WAL doing their best.</em><br> <strong>1 POINT TO ORACLE!</strong></p><p>Moving on to replication. Oracle boasts <strong>RAC with GTT</strong> — meanwhile, PostgreSQL scrambles around with <strong>Patroni and PGTT</strong>, but the <em>bloating</em> makes it feel like a lost cause.<br> <strong>1 POINT TO ORACLE!</strong></p><p>Now, datatype quirks. Oracle SQL doesn’t <em>believe</em> in Boolean (because who needs true or false anyway?), but PostgreSQL and PL/SQL do.<br> PostgreSQL also throws in JSON support, because why not?<br> <strong>1 POINT TO ORACLE!</strong> (because JSON is apparently <em>optional</em> in this contest )</p><p>UUIDs? PostgreSQL has them. Oracle? <em>Just use sequences, peasant.</em></p><p>Execution optimization? Oracle lets you <strong>lock in</strong> an execution plan like a pro with <strong>SQL Plan Management</strong>. PostgreSQL? Some <em>gentle suggestions</em> with hints.<br> <strong>1 POINT TO ORACLE!</strong></p><p>Exception handling? Oracle is flexible, but don’t count on SAVEPOINTS to be your hero. In fact, SAVEPOINTS bring <strong>more locks and SLRU degradation</strong>, because PostgreSQL loves a little chaos.<br> <strong>1 POINT TO ORACLE!</strong></p><h3><em>Final Score:</em></h3><p>Oracle: A smug lead<br>PostgreSQL: Holding on for dear life</p><blockquote>At the end of the day, it’s all about what you can afford — freedom, or a powerful overlord.</blockquote><p>Of course, this is just my take on things, and I’m sure there are plenty of nuances I may have missed (or <em>horribly</em> misunderstood ). If you’ve got insights, counterpoints, or just want to passionately defend your favorite database, let’s talk!</p><p>Drop your thoughts in the comments — let’s make this a battle <em>everyone</em> can learn from. And if I’ve made any technical blunders, consider this an open invitation to correct me (gently, please ).</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5086499c06da" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Color Palette Analyzer]]></title>
            <link>https://medium.com/@plalindia01/color-palette-analyzer-94ae34826a06?source=rss-2a468d39b08f------2</link>
            <guid isPermaLink="false">https://medium.com/p/94ae34826a06</guid>
            <category><![CDATA[genai]]></category>
            <category><![CDATA[fastapi]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[llama-2]]></category>
            <dc:creator><![CDATA[prachi lal]]></dc:creator>
            <pubDate>Tue, 03 Dec 2024 10:09:21 GMT</pubDate>
            <atom:updated>2024-12-03T10:10:49.192Z</atom:updated>
            <content:encoded><![CDATA[<p>This project leverages the power of AI to analyze the skin tone from an uploaded image and generates a curated color palette tailored to the individual’s complexion. By utilizing the Replicate API and the Colorthief library, this application provides a unique and personalized color recommendation experience.</p><h4>Installation</h4><p><strong>A. Cloning the repository:</strong></p><pre>git clone https://github.com/Prachi-Lal/color-palette-analyzer.git</pre><p><strong>B. Create a Virtual Environment:</strong></p><p>This step is completely optional. I have built the project compatible with Python 3.12.</p><pre>python -m venv venv</pre><p><strong>C. Activating the Virtual Environment:</strong></p><p>Ignore if step B is ignored.</p><pre>venv\Scripts\activate</pre><p><strong>D. Install dependencies:</strong></p><pre>pip install -r requirements.txt</pre><p>E. <strong>Set Up Replicate API Token:</strong></p><ul><li>Create a Replicate account and obtain an API token.</li><li>Set the REPLICATE_API_TOKEN environment variable:</li></ul><pre>os.environ[“REPLICATE_API_TOKEN”] = your_api_tokenp</pre><h4>Usage:</h4><p><strong>A. Run the Application:</strong></p><pre>uvicorn app:app --reload</pre><p><strong>B. Access the API:</strong></p><p><strong>Upload an Image:</strong> Send a POST request to the /uploadfile/ endpoint with the image file. Can be done using the Swagger UI.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HZV8mABdsAK2sQR6623DKg.png" /></figure><p><strong>C. Sample Output:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QcstNQTiN-FlpAu6T2lyJQ.png" /></figure><h4><strong>Key Technologies:</strong></h4><ul><li><strong>FastAPI:</strong> A high-performance web framework for building APIs.</li><li><strong>Colorthief:</strong> A Python library for color extraction from images.</li><li><strong>Replicate:</strong> A platform for running AI models.</li><li><strong>LLaMA-2–70B-Chat:</strong> A powerful language model for generating text.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=94ae34826a06" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why you should attend GDG DevFest at your City]]></title>
            <link>https://medium.com/@plalindia01/why-you-should-attend-gdg-devfest-at-your-city-ecb237cc71c1?source=rss-2a468d39b08f------2</link>
            <guid isPermaLink="false">https://medium.com/p/ecb237cc71c1</guid>
            <category><![CDATA[gdgpune]]></category>
            <category><![CDATA[gdsc]]></category>
            <category><![CDATA[gdsc-lead]]></category>
            <category><![CDATA[devfest]]></category>
            <category><![CDATA[gdg]]></category>
            <dc:creator><![CDATA[prachi lal]]></dc:creator>
            <pubDate>Fri, 15 Nov 2024 05:01:12 GMT</pubDate>
            <atom:updated>2024-11-15T05:01:12.381Z</atom:updated>
            <content:encoded><![CDATA[<p>Recently I attended DevFest Pune 2023. It was a blast! If you are someone who hasn’t attended a <a href="https://www.linkedin.com/company/gdgpune/">GDG</a> DevFest till now, this post is for you. I will tell you exactly why you need to be on the lookout for the tickets. Also I will be tagging LinkedIn accounts for as many people I can so that it’s easier for y’all to stalk.</p><p>First of all kudos to the organizers- <a href="https://www.linkedin.com/in/mahaveer-muttha/">Mr. Mahaveer Mutha</a>, <a href="https://www.linkedin.com/in/namratamore/">Ms. Namrata More</a>, <a href="https://www.linkedin.com/in/swapniljathar/">Mr. Swapnil Jathar</a> and <a href="https://www.linkedin.com/in/pranoti-nandurkar-170021b/">Ms. Pranoti Nandurkar</a> for organizing an event at such a huge scale with such finesse and competence.</p><p>One could understand the craze amongst students about this event as the student tickets were out within 6 hours of announcement. I was lucky enough to get a ticket somehow. The fest had 27 different talks and 3 panel discussions distributed in 4 different sections.</p><p>The event started with <a href="https://www.linkedin.com/in/manoranjanpadhy/">Mr. Manoranjan Padhy</a> as the welcome keynote speaker talking about Responsible AI and then <a href="https://www.linkedin.com/in/vishalvirani/">Mr. Vishal Virani</a> discussing how AI has impacted various aspects of our lives. We were also introduced to <a href="https://www.linkedin.com/company/dhiwise/">DhiWise</a>, a programming platform where one can convert their designs into developer-friendly code., followed by live demos by their team.</p><p>I got to attend an interesting talk featuring <a href="https://www.linkedin.com/in/ajeygore/">Mr. Ajey Gore</a> and <a href="https://www.linkedin.com/in/gautamrege/">Mr. Gautam Rege</a> on how advancements in AI has caused disruptions in the industry and our day to day life. Some important words that I remember from the talk- uncomfortable, driver and redundant nor replaced. All the people who attended can testify that these words signify a rich, interesting context. I can undoubtedly say that out of all the talks, this one was my absolute favorite and I wanted the talk to last more.</p><p>I also got a chance to talk to <a href="https://www.linkedin.com/in/dr-sonal-sushil-modi-ph-d-in-management-cplp%C2%AE-b03385b/">Dr. Sonal Modi</a> about women in tech and how one’s gender shouldn’t be a hinderance in their path.</p><p><a href="https://www.linkedin.com/in/mathangisri/">Mrs. Mathangi Sri Ramachandran</a> shed some light on how Yubi deals with loads of documentation with much human intervention. People at Yubi are trying to build tools which deal with a lot of sensitive and important paperwork.</p><p>Then we had the delicious lunch catered by none other than J. W. Mariott, Pune. There were a lot of fun activities planned- contests and fun games to win exciting goodies. I also ended up giving me resume out to representatives of different companies (yes, I carried a hard copy of my resume) and received some constructive suggestions as well.</p><p>Even in the second half we had a ton of talks and discussions planned, like basics of LLMs by <a href="https://www.linkedin.com/in/yogeshkulkarni/">Mr. Yogesh Kulkarni</a> and insights on project management by <a href="https://www.linkedin.com/in/ayushj/">Mr. Ayush Jain</a>. One post is not enough to summarize all the talk and sessions we had. I also got to reconnect with all the GDSC Leads from sister colleges and started rolling the wheels on something else (we’ll be announcing soon). Last but not the least- the panel discussion on Diversity and Inclusion rounded up the day on the perfect note.</p><p>I went to the DevFest alone, but it’s a DevFest. One can talk to anyone and everyone. It was indeed a day spent well.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ecb237cc71c1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Foundation Model Guardrails]]></title>
            <link>https://medium.com/@plalindia01/foundation-model-guardrails-94c5608c33a7?source=rss-2a468d39b08f------2</link>
            <guid isPermaLink="false">https://medium.com/p/94c5608c33a7</guid>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[aws-bedrock]]></category>
            <category><![CDATA[llm-guardrails]]></category>
            <category><![CDATA[aws]]></category>
            <dc:creator><![CDATA[prachi lal]]></dc:creator>
            <pubDate>Sun, 10 Nov 2024 07:46:46 GMT</pubDate>
            <atom:updated>2024-11-10T07:46:46.965Z</atom:updated>
            <content:encoded><![CDATA[<p>Used to control the interaction between the users and the foundation models.</p><p>The developer can set certain boundaries with respect to what type of content is considered harmful and undesirable and how to filter these out.</p><p>One can also program it to remove any personally identifiable information (PII), therefore, enhancing privacy.</p><p>Guardrails also come in handy to reduce hallucinations.</p><p><em>In the following passages, I will be setting up a test guardrail in the AWS Bedrock architecture.</em></p><p>Step #1: Initialization</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tfzejAY1o4hX_HLDikj2_g.png" /></figure><p>Different or same messages can be used for both blocked response or blocked message respectively.</p><p>Step #2: Built-in filters</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2qWjDteeDHS_zYiCktR4bA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qxgCAR0zZITe4AE1tzBz8g.png" /></figure><p>This section allows to set some built-in features to filter out harmful categories. There is also a provision of input tagging to avoid prompt attacks.</p><p>Step #3: Adding specific denied topics</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cNr-AINCFl2sWqIjox_IiA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/905/1*gm_F5dyukJ4KMl062olmmA.png" /></figure><p>Step #4: Adding a filter for a built-in list of globally accepted profane words and other custom words and phrases</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pVFzndIEZrN6EqRRWgnBiQ.png" /></figure><p>These words and phrases can be entered by various source- manually using the GUI, uploading a .txt or .csv files or via a S3 object.</p><p>Step #5: Filtering sensitive information</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kTTpJD2Hz0Su7lwQ78vQdQ.png" /></figure><p>The PII can be either blocked or masked. Sensitive information can also be identified for filtering using Regex expressions.</p><p>Step #6: Grounding and Relevance Check</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Lgyfa1USi1UWkzRSArMIqw.png" /></figure><p>Step #7: Review the guardrail and create</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AgienY-N_WFkqlzJaHv2Zg.png" /></figure><h4><strong>Testing</strong></h4><p>AWS Bedrock has a provision for testing the guardrail using custom prompts to a selected model.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-ZM_UbLgLydNk8-kJJTOrg.png" /></figure><p>That’s it for this article. Do let me know if it was fun understanding Guardrails and how I can make this learning experience better!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=94c5608c33a7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Telecom Churn ESA- A Case Study]]></title>
            <link>https://medium.com/@plalindia01/telecom-churn-esa-a-case-study-0eb798e718b1?source=rss-2a468d39b08f------2</link>
            <guid isPermaLink="false">https://medium.com/p/0eb798e718b1</guid>
            <category><![CDATA[eda]]></category>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[python]]></category>
            <dc:creator><![CDATA[prachi lal]]></dc:creator>
            <pubDate>Mon, 21 Oct 2024 23:53:05 GMT</pubDate>
            <atom:updated>2024-10-21T23:53:05.611Z</atom:updated>
            <content:encoded><![CDATA[<h3>Telecom Churn EDA- A Case Study</h3><p>The churn rate, also known as the rate of attrition or customer churn, is the rate at which customers stop doing business with an entity.</p><p>Through proper analysis, businesses can derive insights and understand the key drivers of churn related to them at both organizational and customer levels.</p><blockquote><em>Disclaimer: This is not my original work. This is rather an attempt at duplication for learning purposes.</em></blockquote><p>I will be using this <a href="https://github.com/Prachi-Lal/telecom_churn_pred/blob/main/WA_Fn-UseC_-Telco-Customer-Churn.csv">dataset</a> for the case study.</p><p>Let’s get started with the Exploratory Data Analysis(EDA) to understand the peculiarities of our dataset.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tmVOhxh7dFKjOMoTSqiiSA.png" /></figure><ul><li>75% customers have tenure less than 55 months</li><li>Average Monthly charges are USD 64.76 whereas 25% customers pay more than USD 89.85 per month</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lXjVDMQJcHc4MgQ7jkS85Q.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1UWat_wO9Zv1NYYVjoHN1Q.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wUboC__-kjMyP4pg1tg5sg.png" /></figure><ul><li>Data is highly imbalanced, ratio = 73:27</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4YbgI4rTZN1aOIKAZ5cQZA.png" /></figure><ul><li>Here, we don&#39;t have any missing data. Phew!</li></ul><p>There are some eccentricities in our dataset, which can be cured using simple data cleaning methods.</p><ul><li>I am very scared of tweaking the original data, hence a copy of it has to be made.</li><li>Total charges should be of numeric data type.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PZeP5Fdur_Iqq58WNwXFGQ.png" /></figure><ul><li>There are 11 missing values in the “TotalCharges” column.</li><li>Since the percent of these records is very low (0.15%), it is viable to ignore them for future processing.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_ofw2zfdzvvMRFS6bwmgDQ.png" /></figure><ul><li>Tenure can be put into bins.</li><li>Remove unnecessary columns.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-FdSQxHrPXahalCPQQSocA.png" /></figure><p>Let’s proceed to data exploration and visualization!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wmKiFcb2FL4FJAYJF9GjLg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_1xvnsHW8TAU3ProEwEHBQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*B4ctuwOZ1OEJlDAMmCW29w.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hB2CFbxBgIlStIJA9dqT4A.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*brFQWSu5KLqWJ8o1JdnAuw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VW8C05pYV9SWLOpZGlSNZQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-SrV1YTynLi_xjKt-Vp7UQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8K8tokso0BdU7I6eXbWPsQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YwJl5Hkcl9uxddq4wHArgA.png" /></figure><ul><li>Convert the target variable “Churn” in a binary numeric variable.</li><li>Converting all the categorical variable into dummy variables.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BIjsy-X8ocUx-QK3KljZdA.png" /></figure><p><strong>Relationship between Monthly Charges and Total Charges</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/832/1*EZW-5Iu4EMO9CHDkdDBxxQ.png" /></figure><p>Total Charges increase as Monthly Charges increase</p><p><strong>Churn by Monthly Charges and Total Charges</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/886/1*-0YzEBxYv6iec2iAQByO0Q.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/692/1*ai6tjNEljcA58XXV8xIzBw.png" /></figure><p>Churn is high when Monthly Charges are high</p><p><strong>Total Charges and Churn</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/884/1*yB_cnaE9xkT7gVS780_lpg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/710/1*JCSgKoCKhep_a5VebzM1oA.png" /></figure><p><strong>Higher churn at a lower Total charge.</strong> Well, that’s surprising!</p><p>Correlation of all predictors with Churn.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7bluD4K0plN4Y0Ic11ToVw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/990/1*q5kSfFKYBcy0qJFAYQtR6Q.png" /></figure><p>HIGH Churn seen in case of Month to month contracts, No online security, No Tech support, First year of subscription and Fiber Optics Internet</p><p>LOW Churn is seen in case of Long term contracts, Subscriptions without internet service and the customers engaged for 5+ years</p><p>Factors like Gender, Availability of PhoneService and # of multiple lines have almost NO impact on Churn</p><p>Let’s observe some more bivariate analysis!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jk5yz1MVYEKEc7eVm13E8A.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tzU_FT-fgpA6gBpA2bPv9Q.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*z3CUEcTuxb6iFkTwSQ6wXA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fMYtfQpl_YPnVUNgZ-0GXg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XD0yeAQXS5Ed8DD7e0IOng.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XfUn2Fyqkc14Ey77zLvb4w.png" /></figure><p><strong>Concluding remarks:<br></strong>Electronic check medium are the highest churners<br>Contract Type — Monthly customers are more likely to churn because of no contract terms, as they are free to go customers.<br>No Online security, No Tech Support category are high churners<br>Non senior Citizens are high churners</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/889/1*gtt5L7tzD9uTpJVge-n6Fw.png" /></figure><p>This cured csv can be found <a href="https://github.com/Prachi-Lal/telecom_churn_pred/blob/main/tel_churn.csv">here</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0eb798e718b1" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>