<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Ujjwal Kumar</title>
    <description>The latest articles on DEV Community by Ujjwal Kumar (@uxzwal).</description>
    <link>https://dev.to/uxzwal</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F1322214%2F50d4e5b3-a521-4d53-a16a-2ef96501af18.png</url>
      <title>DEV Community: Ujjwal Kumar</title>
      <link>https://dev.to/uxzwal</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/uxzwal"/>
    <language>en</language>
    <item>
      <title>WARDEN: Hunting the Ghost in the Machine with an AI-Powered Self-Healing DevOps Agent</title>
      <dc:creator>Ujjwal Kumar</dc:creator>
      <pubDate>Fri, 30 Oct 2026 05:44:04 +0000</pubDate>
      <link>https://dev.to/uxzwal/warden-hunting-the-ghost-in-the-machine-with-an-ai-powered-self-healing-devops-agent-45b9</link>
      <guid>https://dev.to/uxzwal/warden-hunting-the-ghost-in-the-machine-with-an-ai-powered-self-healing-devops-agent-45b9</guid>
      <description>&lt;p&gt;&lt;em&gt;This is a submission for the &lt;a href="https://dev.to/challenges/hacktoberfest-week4-2026-10-26"&gt;Hacktoberfest Open-Source AI Challenge Week 4: Ghost in the Machine&lt;/a&gt;&lt;/em&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  What I Built
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;WARDEN is an AI-powered self-healing DevOps agent that detects service failures, investigates their root causes, proposes safe remediation actions, and verifies whether the system has actually recovered.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Every production system has its ghosts.&lt;/p&gt;

&lt;p&gt;A sudden traffic spike. An exhausted connection pool. Unexpected HTTP 503 errors. A service that appears healthy but silently struggles under concurrent load.&lt;/p&gt;

&lt;p&gt;These failures can be difficult to diagnose. Engineers often spend valuable time switching between logs, metrics, configuration files, and monitoring dashboards.&lt;/p&gt;

&lt;p&gt;I wanted to explore a different approach:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;What if an AI agent could investigate incidents and help resolve them while keeping safety and human approval at the center?&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;That idea became WARDEN.&lt;/p&gt;

&lt;p&gt;WARDEN follows a four-stage incident response workflow:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Detect → Diagnose → Remediate → Verify&lt;/strong&gt;&lt;/p&gt;

&lt;h3&gt;
  
  
  Key Features
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Incident Detection:&lt;/strong&gt; Monitors service health, telemetry, logs, latency, and HTTP failure rates.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;AI-Powered Diagnosis:&lt;/strong&gt; Analyzes incident information and identifies potential root causes.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Safe Remediation:&lt;/strong&gt; Restricts proposed actions to a predefined allowlist with strict parameter boundaries.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Human-in-the-Loop Approval:&lt;/strong&gt; Requires operator authorization before applying configuration changes.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Empirical Recovery Verification:&lt;/strong&gt; Uses concurrent HTTP requests to determine whether the service has genuinely recovered.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Audit Trail:&lt;/strong&gt; Maintains structured records of incident and remediation events.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  The Ghost in the Machine
&lt;/h3&gt;

&lt;p&gt;For the demonstration, WARDEN manages a FastAPI-based payment-service simulator.&lt;/p&gt;

&lt;p&gt;I can deliberately introduce connection-pool exhaustion, causing the service to return HTTP 503 errors under concurrent traffic.&lt;/p&gt;

&lt;p&gt;WARDEN then:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Detects the service degradation.&lt;/li&gt;
&lt;li&gt;Analyzes metrics and logs to identify connection-pool starvation.&lt;/li&gt;
&lt;li&gt;Proposes an increase to the connection-pool limit.&lt;/li&gt;
&lt;li&gt;Waits for explicit operator approval.&lt;/li&gt;
&lt;li&gt;Applies the permitted configuration change.&lt;/li&gt;
&lt;li&gt;Runs another concurrent HTTP verification to measure recovery.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;The important part is that WARDEN does not simply announce that an incident is fixed because a configuration changed.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;It verifies recovery using actual HTTP responses.&lt;/strong&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Demo
&lt;/h2&gt;

&lt;p&gt;Watch WARDEN in action:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://youtu.be/rc6QmCs6KbI" rel="noopener noreferrer"&gt;▶️ Watch the WARDEN Project Demo on YouTube&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The demonstration showcases the project's incident-response workflow and recovery capabilities.&lt;/p&gt;

&lt;h2&gt;
  
  
  Code
&lt;/h2&gt;

&lt;p&gt;The project is open source and available on GitHub:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://github.com/uxzwal/Warden-AI-Powered-Self-Healing-DevOps-Agent" rel="noopener noreferrer"&gt;WARDEN - AI-Powered Self-Healing DevOps Agent&lt;/a&gt;&lt;/p&gt;

&lt;h3&gt;
  
  
  Technology Stack
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;Python&lt;/li&gt;
&lt;li&gt;FastAPI&lt;/li&gt;
&lt;li&gt;Streamlit&lt;/li&gt;
&lt;li&gt;Nebius Token Factory&lt;/li&gt;
&lt;li&gt;OpenAI-compatible SDK&lt;/li&gt;
&lt;li&gt;Pydantic&lt;/li&gt;
&lt;li&gt;HTTPX&lt;/li&gt;
&lt;li&gt;Pytest&lt;/li&gt;
&lt;li&gt;Tavily (optional technical research integration)&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  Project Architecture
&lt;/h3&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;WARDEN/
│
├── warden/
│   ├── app.py
│   ├── payment_service.py
│   ├── agent.py
│   ├── nebius_client.py
│   ├── tavily_client.py
│   ├── verifier.py
│   ├── remediation.py
│   ├── schemas.py
│   └── audit.py
│
├── tests/
├── requirements.txt
├── pytest.ini
└── README.md
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The repository includes automated tests for service behavior, remediation safety, agent output validation, and empirical recovery verification.&lt;/p&gt;

&lt;h2&gt;
  
  
  How I Built It
&lt;/h2&gt;

&lt;p&gt;WARDEN combines an AI reasoning layer with deterministic execution controls.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. AI Reasoning
&lt;/h3&gt;

&lt;p&gt;The agent architecture supports open-weight models through Nebius Token Factory's OpenAI-compatible API.&lt;/p&gt;

&lt;p&gt;The documented model configurations include:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Meta-Llama-3.1-8B-Instruct for lower-latency diagnostic tasks.&lt;/li&gt;
&lt;li&gt;Meta-Llama-3.1-70B-Instruct for more complex incident analysis.&lt;/li&gt;
&lt;li&gt;DeepSeek-R1 as an alternative reasoning model configuration.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The architecture separates fast triage from deeper reasoning so that model selection can match the complexity of an incident.&lt;/p&gt;

&lt;p&gt;A deterministic Demo Mode is also available, allowing the incident-response workflow to operate without API credentials.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. Safety Before Automation
&lt;/h3&gt;

&lt;p&gt;Giving an AI agent unrestricted access to production infrastructure would introduce serious operational risks.&lt;/p&gt;

&lt;p&gt;WARDEN takes a constrained approach.&lt;/p&gt;

&lt;p&gt;The agent cannot execute arbitrary shell commands, Python code, SQL, or unrestricted operations.&lt;/p&gt;

&lt;p&gt;Instead, remediation is limited to explicitly permitted actions:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;&lt;code&gt;increase_pool_limit&lt;/code&gt;&lt;/li&gt;
&lt;li&gt;&lt;code&gt;reset_service_config&lt;/code&gt;&lt;/li&gt;
&lt;li&gt;&lt;code&gt;no_op&lt;/code&gt;&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Numeric boundaries are enforced through deterministic validation, and configuration changes require explicit operator approval.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. Real Recovery Verification
&lt;/h3&gt;

&lt;p&gt;WARDEN uses HTTPX and asynchronous requests to evaluate service behavior under concurrent load.&lt;/p&gt;

&lt;p&gt;It measures actual HTTP responses and performance indicators instead of trusting the agent's claims.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;A proposed fix is not the same thing as a verified recovery.&lt;/strong&gt;&lt;/p&gt;

&lt;h3&gt;
  
  
  4. Testing
&lt;/h3&gt;

&lt;p&gt;The project includes 18 automated tests covering:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Service health and concurrency behavior.&lt;/li&gt;
&lt;li&gt;HTTP 503 responses under overload.&lt;/li&gt;
&lt;li&gt;Administrative authentication.&lt;/li&gt;
&lt;li&gt;Remediation allowlists and parameter boundaries.&lt;/li&gt;
&lt;li&gt;Human approval enforcement.&lt;/li&gt;
&lt;li&gt;Structured diagnostic validation.&lt;/li&gt;
&lt;li&gt;Recovery verification.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Why Does Open Innovation Matter?
&lt;/h2&gt;

&lt;p&gt;Open innovation matters because infrastructure automation should not depend on a single closed model or an opaque decision-making system.&lt;/p&gt;

&lt;p&gt;Using open-weight model options through an inference platform such as Nebius Token Factory provides flexibility in model selection while allowing developers to retain control over the surrounding application architecture.&lt;/p&gt;

&lt;p&gt;For WARDEN, the reasoning model is only one component of the system.&lt;/p&gt;

&lt;p&gt;The application also contains independently testable safety validators, structured schemas, deterministic remediation rules, and recovery verification.&lt;/p&gt;

&lt;p&gt;This separation makes it possible to experiment with different models without giving them unrestricted control over infrastructure operations.&lt;/p&gt;

&lt;p&gt;Open innovation also makes it easier for developers to inspect, adapt, test, and extend the system according to their own operational requirements.&lt;/p&gt;

&lt;h2&gt;
  
  
  My Agent Session
&lt;/h2&gt;

&lt;p&gt;WARDEN's incident lifecycle can be demonstrated through its dashboard:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Incident → Diagnosis → Approval → Remediation → Verification&lt;/strong&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  The Spirit of It
&lt;/h2&gt;

&lt;p&gt;The most interesting part of building WARDEN was realizing that an AI agent should not be judged only by how intelligently it explains a problem.&lt;/p&gt;

&lt;p&gt;It should also be judged by what happens after its recommendation.&lt;/p&gt;

&lt;p&gt;Does the service actually recover?&lt;/p&gt;

&lt;p&gt;Can the action be trusted?&lt;/p&gt;

&lt;p&gt;Can an engineer understand what changed?&lt;/p&gt;

&lt;p&gt;And can the system prove that the incident is resolved?&lt;/p&gt;

&lt;p&gt;I wanted to build something that goes beyond an AI chatbot suggesting commands.&lt;/p&gt;

&lt;p&gt;WARDEN represents an approach where AI reasoning works together with deterministic safety, human oversight, and measurable outcomes.&lt;/p&gt;

&lt;p&gt;The ghost in the machine might be invisible, but its impact is measurable.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;WARDEN: Find the failure. Understand the cause. Fix it safely. Prove the recovery.&lt;/strong&gt;&lt;/p&gt;

</description>
      <category>devchallenge</category>
      <category>hf26challenge</category>
      <category>ai</category>
      <category>devops</category>
    </item>
    <item>
      <title>Menu-Sahayak: Instant Allergen Insights for Indian Restaurant Menus (Powered by Llama 3.3)</title>
      <dc:creator>Ujjwal Kumar</dc:creator>
      <pubDate>Sun, 18 Oct 2026 04:40:17 +0000</pubDate>
      <link>https://dev.to/uxzwal/menu-sahayak-instant-allergen-insights-for-indian-restaurant-menus-powered-by-llama-33-3mla</link>
      <guid>https://dev.to/uxzwal/menu-sahayak-instant-allergen-insights-for-indian-restaurant-menus-powered-by-llama-33-3mla</guid>
      <description>&lt;p&gt;&lt;em&gt;This is a submission for the &lt;a href="https://dev.to/challenges/hacktoberfest-week2-2026-10-12"&gt;Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle&lt;/a&gt;&lt;/em&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  What I Built
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;Menu-Sahayak&lt;/strong&gt; is a lightweight web app that analyzes Indian restaurant menus and flags likely allergens per dish. &lt;/p&gt;

&lt;p&gt;Tourists with food allergies often struggle to read Hindi or regional-language menus. Generic translation tools will tell you what "Chole Bhature" means, but they won't tell you it contains gluten and dairy. Menu-Sahayak bridges that gap. Paste a menu, and instantly get dish names alongside clear allergen tags (dairy, nuts, gluten, eggs, soy) and a practical safety note for each item.&lt;/p&gt;

&lt;p&gt;It solves a very real, high-stakes problem for travelers and locals alike, making dining out safer and stress-free.&lt;/p&gt;

&lt;h2&gt;
  
  
  Demo
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F4243yfk5nts2dc0fw4j0.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F4243yfk5nts2dc0fw4j0.png" alt=" " width="800" height="533"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Here is a quick look at the interface:&lt;br&gt;
&lt;em&gt;(Apna screenshot yahan upload karein — DEV editor mein image icon par click karke)&lt;/em&gt;&lt;/p&gt;
&lt;h2&gt;
  
  
  Code
&lt;/h2&gt;


&lt;div class="ltag-github-readme-tag"&gt;
  &lt;div class="readme-overview"&gt;
    &lt;h2&gt;
      &lt;img src="https://assets.dev.to/assets/github-logo-5a155e1f9a670af7944dd5e12375bc76ed542ea80224905ecaf878b9157cdefc.svg" alt="GitHub logo"&gt;
      &lt;a href="https://github.com/uxzwal" rel="noopener noreferrer"&gt;
        uxzwal
      &lt;/a&gt; / &lt;a href="https://github.com/uxzwal/Menu-Sahayak" rel="noopener noreferrer"&gt;
        Menu-Sahayak
      &lt;/a&gt;
    &lt;/h2&gt;
    &lt;h3&gt;
      
    &lt;/h3&gt;
  &lt;/div&gt;
  &lt;div class="ltag-github-body"&gt;
    
&lt;div id="readme" class="md"&gt;&lt;div class="markdown-heading"&gt;
&lt;h1 class="heading-element"&gt;Allergen Lens — Menu Allergen Checker for Indian Restaurants&lt;/h1&gt;
&lt;/div&gt;
&lt;p&gt;A lightweight Gradio app that analyzes Indian restaurant menu items and flags &lt;strong&gt;likely allergens&lt;/strong&gt; per dish.&lt;/p&gt;
&lt;p&gt;Powered by &lt;strong&gt;Llama 3.3 70B (open-weight)&lt;/strong&gt; via &lt;a href="https://groq.com" rel="nofollow noopener noreferrer"&gt;Groq&lt;/a&gt;.&lt;/p&gt;
&lt;div class="markdown-heading"&gt;
&lt;h2 class="heading-element"&gt;What it does&lt;/h2&gt;
&lt;/div&gt;
&lt;p&gt;Paste menu items (one dish per line), and the app returns a numbered list with:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;dish name&lt;/li&gt;
&lt;li&gt;likely allergens from: dairy, nuts, gluten, eggs, soy&lt;/li&gt;
&lt;li&gt;one concise safety note&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;The model is instructed to avoid unsupported certainty: allergens are marked as likely/possible based on common preparation, and users should verify with restaurant staff for severe allergies.&lt;/p&gt;
&lt;div class="markdown-heading"&gt;
&lt;h2 class="heading-element"&gt;Why this is useful&lt;/h2&gt;
&lt;/div&gt;
&lt;p&gt;Tourists with allergies may struggle to read Hindi or regional-language menus. This tool provides quick, practical allergen guidance dish-by-dish.&lt;/p&gt;
&lt;div class="markdown-heading"&gt;
&lt;h2 class="heading-element"&gt;Setup&lt;/h2&gt;

&lt;/div&gt;
&lt;ol&gt;
&lt;li&gt;Clone this repository.&lt;/li&gt;
&lt;li&gt;Install dependencies:
&lt;div class="highlight highlight-source-shell notranslate position-relative overflow-auto js-code-highlight"&gt;
&lt;pre&gt;pip install -r menu-sahayak/requirements.txt&lt;/pre&gt;

&lt;/div&gt;
&lt;/li&gt;
&lt;li&gt;Create env file:
&lt;div class="highlight highlight-source-shell notranslate position-relative overflow-auto js-code-highlight"&gt;
&lt;pre&gt;cp menu-sahayak/.env.example menu-sahayak/.env&lt;/pre&gt;

&lt;/div&gt;
&lt;/li&gt;
&lt;li&gt;Add your Groq API key in &lt;code&gt;menu-sahayak/.env&lt;/code&gt;:
&lt;div class="highlight highlight-source-shell notranslate position-relative overflow-auto js-code-highlight"&gt;
&lt;pre&gt;GROQ_API_KEY=your_key_here&lt;/pre&gt;

&lt;/div&gt;
&lt;/li&gt;
&lt;li&gt;Run:
&lt;div class="highlight highlight-source-shell notranslate position-relative overflow-auto js-code-highlight"&gt;
&lt;pre&gt;&lt;span class="pl-c1"&gt;cd&lt;/span&gt; menu-sahayak
python app.py&lt;/pre&gt;

&lt;/div&gt;
&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;Get a free Groq…&lt;/p&gt;&lt;/div&gt;
  &lt;/div&gt;
  &lt;div class="gh-btn-container"&gt;&lt;a class="gh-btn" href="https://github.com/uxzwal/Menu-Sahayak" rel="noopener noreferrer"&gt;View on GitHub&lt;/a&gt;&lt;/div&gt;
&lt;/div&gt;


&lt;h2&gt;
  
  
  How I Built It
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Model:&lt;/strong&gt; Llama 3.3 70B (open-weight) served via Groq's fast inference API.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Frontend/UI:&lt;/strong&gt; Python + Gradio (customized theme for a clean, mobile-friendly dashboard feel).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;The Core Logic:&lt;/strong&gt; The app streams the response from the model and uses a highly specific system prompt to force the model into an "Allergen Analyst" persona. Instead of creative writing, it outputs structured, numbered allergen lists.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Error Handling:&lt;/strong&gt; Built-in graceful fallbacks for missing API keys, rate limits, and empty inputs.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Why Does Open Innovation Matter?
&lt;/h2&gt;

&lt;p&gt;I chose &lt;strong&gt;Llama 3.3 70B (open-weight)&lt;/strong&gt; over a closed API because:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Cost-Effective for a Side Hustle:&lt;/strong&gt; Groq's generous free tier makes this app &lt;strong&gt;₹0 to run&lt;/strong&gt;. A closed model with similar reasoning capabilities would charge per token, eating into the thin margins of a side project.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Model Swap Freedom:&lt;/strong&gt; Because the logic is built around an open-weight model, I am not locked into one provider. If I want to swap Llama for Qwen or Gemma tomorrow to improve allergy detection, it's a one-line change.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Transparency and Control:&lt;/strong&gt; Open weights mean I know exactly what is processing the user's input. For an app dealing with &lt;em&gt;health and dietary safety&lt;/em&gt;, this transparency is crucial.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;The open pieces aren't just a buzzword here — the entire product is built around an accessible, open-weight model that I can control, prompt, and eventually self-host if needed.&lt;/p&gt;

&lt;h2&gt;
  
  
  Prize Categories
&lt;/h2&gt;

&lt;p&gt;Overall Side Hustle Challenge Entry.&lt;/p&gt;

&lt;h2&gt;
  
  
  #devchallenge #hf26challenge #opensource #ai #llama #python #sidehustle
&lt;/h2&gt;

</description>
      <category>devchallenge</category>
      <category>hf26challenge</category>
      <category>opensource</category>
      <category>ai</category>
    </item>
    <item>
      <title>🚀 Built &amp; deployed my new 3D DevOps Portfolio

Tech Stack:
⚛️ React + Vite
🎨 Tailwind CSS
🧊 Three.js
🎬 Framer Motion
🔥 Supabase
🐳 Docker + GitHub Actions

Features:
• Interactive 3D UI
• Modern animations
• Responsive design
• DevOps-inspired archite</title>
      <dc:creator>Ujjwal Kumar</dc:creator>
      <pubDate>Wed, 27 May 2026 10:11:08 +0000</pubDate>
      <link>https://dev.to/uxzwal/built-deployed-my-new-3d-devops-portfolio-tech-stack-react-vite-tailwind-css-15n8</link>
      <guid>https://dev.to/uxzwal/built-deployed-my-new-3d-devops-portfolio-tech-stack-react-vite-tailwind-css-15n8</guid>
      <description></description>
      <category>devops</category>
      <category>frontend</category>
      <category>react</category>
      <category>showdev</category>
    </item>
    <item>
      <title>Happy Holi</title>
      <dc:creator>Ujjwal Kumar</dc:creator>
      <pubDate>Tue, 19 Mar 2024 11:39:52 +0000</pubDate>
      <link>https://dev.to/uxzwal/happy-holi-59p0</link>
      <guid>https://dev.to/uxzwal/happy-holi-59p0</guid>
      <description>&lt;p&gt;Check out this Pen I made!&lt;/p&gt;

&lt;p&gt;&lt;iframe height="600" src="https://codepen.io/you_jjwal/embed/poBRLNj?height=600&amp;amp;default-tab=result&amp;amp;embed-version=2"&gt;
&lt;/iframe&gt;
&lt;/p&gt;

</description>
      <category>codepen</category>
    </item>
    <item>
      <title>Neon Name Tagline</title>
      <dc:creator>Ujjwal Kumar</dc:creator>
      <pubDate>Mon, 18 Mar 2024 14:10:14 +0000</pubDate>
      <link>https://dev.to/uxzwal/neon-name-tagline-1nal</link>
      <guid>https://dev.to/uxzwal/neon-name-tagline-1nal</guid>
      <description>&lt;p&gt;Check out this Pen I made!&lt;/p&gt;

&lt;p&gt;&lt;iframe height="600" src="https://codepen.io/you_jjwal/embed/PogWqKe?height=600&amp;amp;default-tab=result&amp;amp;embed-version=2"&gt;
&lt;/iframe&gt;
&lt;/p&gt;

</description>
      <category>codepen</category>
    </item>
    <item>
      <title>My Pen on CodePen</title>
      <dc:creator>Ujjwal Kumar</dc:creator>
      <pubDate>Mon, 18 Mar 2024 13:57:10 +0000</pubDate>
      <link>https://dev.to/uxzwal/my-pen-on-codepen-186</link>
      <guid>https://dev.to/uxzwal/my-pen-on-codepen-186</guid>
      <description>&lt;p&gt;Check out this Pen I made!&lt;/p&gt;

&lt;p&gt;&lt;iframe height="600" src="https://codepen.io/you_jjwal/embed/OJGWVyz?height=600&amp;amp;default-tab=result&amp;amp;embed-version=2"&gt;
&lt;/iframe&gt;
&lt;/p&gt;

</description>
      <category>codepen</category>
    </item>
    <item>
      <title>Bad Panda</title>
      <dc:creator>Ujjwal Kumar</dc:creator>
      <pubDate>Tue, 05 Mar 2024 13:23:30 +0000</pubDate>
      <link>https://dev.to/uxzwal/bad-panda-2dmj</link>
      <guid>https://dev.to/uxzwal/bad-panda-2dmj</guid>
      <description>&lt;p&gt;Check out this Pen I made!&lt;/p&gt;

&lt;p&gt;&lt;iframe height="600" src="https://codepen.io/you_jjwal/embed/poBJZaM?height=600&amp;amp;default-tab=result&amp;amp;embed-version=2"&gt;
&lt;/iframe&gt;
&lt;/p&gt;

</description>
      <category>codepen</category>
    </item>
    <item>
      <title>Calculator</title>
      <dc:creator>Ujjwal Kumar</dc:creator>
      <pubDate>Tue, 05 Mar 2024 13:21:19 +0000</pubDate>
      <link>https://dev.to/uxzwal/calculator-2pcd</link>
      <guid>https://dev.to/uxzwal/calculator-2pcd</guid>
      <description>&lt;p&gt;Check out this Pen I made!&lt;/p&gt;

&lt;p&gt;&lt;iframe height="600" src="https://codepen.io/you_jjwal/embed/bGJdjQJ?height=600&amp;amp;default-tab=result&amp;amp;embed-version=2"&gt;
&lt;/iframe&gt;
&lt;/p&gt;

</description>
      <category>codepen</category>
    </item>
    <item>
      <title>Battery Animation</title>
      <dc:creator>Ujjwal Kumar</dc:creator>
      <pubDate>Sun, 03 Mar 2024 15:15:26 +0000</pubDate>
      <link>https://dev.to/uxzwal/battrey-animation-1k62</link>
      <guid>https://dev.to/uxzwal/battrey-animation-1k62</guid>
      <description>&lt;p&gt;Check out this Pen I made!&lt;/p&gt;

&lt;p&gt;&lt;iframe height="600" src="https://codepen.io/you_jjwal/embed/GRLgXOR?height=600&amp;amp;default-tab=result&amp;amp;embed-version=2"&gt;
&lt;/iframe&gt;
&lt;/p&gt;

</description>
      <category>codepen</category>
    </item>
  </channel>
</rss>
