DEV Community: Siva Kukkuluri

GaiaOS: An Immersive AI Oracle for Earth Day 🌱 (Auth0 & Gemini)

Siva Kukkuluri — Fri, 17 Apr 2026 15:58:29 +0000

This is a submission for Weekend Challenge: Earth Day Edition

What I Built

For this Earth Day challenge, I set out to solve a core problem in environmental science: we have petabytes of climate data, but a massive bottleneck in actionable synthesis.

I built GaiaOS - The Planetary Intelligence Engine, a "god-tier" interactive command center concept designed for climate scientists, urban planners, and green tech companies. It serves as a unified hub where complex environmental metrics are visualized in real-time, and more importantly, actively parsed by an AI agent to offer sustainable strategies.

Featuring a deeply immersive glassmorphic UI, smooth micro-animations, and a terminal-style developer security gateway, GaiaOS is both visually captivating and functionally robust.

Demo

Exploring the Planetary Engine

Below are the core interfaces of the application. (Please drag and drop the screenshots of the Gemini Oracle and the Auth0 M2M Terminal here before publishing!)

You can pull down the repository and run it locally yourself using the Vite dev server!

Code

07MADARA / GaiaOS-Global-Environmental-Metrics

GaiaOS - The Planetary Intelligence Engine 🌍

A Prototype for the DEV Weekend Challenge: Earth Day Edition

GaiaOS is an advanced, highly immersive environmental data visualization and intelligence dashboard. Designed as a "god-tier" developer prototype, it simulates an enterprise-level platform used by climate scientists, green tech companies, and urban planners to analyze global datasets, track carbon emissions, and run predictive climate models.

(Placeholder representing Earth Intelligence)

🌎 Why We Built This

With Earth Day fast approaching, the challenge prompt "Build for the Planet" inspired us to think about how modern AI and data platforms can actually be leveraged to solve real planetary issues.

The core challenge with environmental action isn't a lack of data—it's a lack of actionable synthesis. There are petabytes of satellite imagery, carbon sensor metrics, and climate documents available. We built GaiaOS to conceptualize a unified command center: a place where massive amounts of data are…

View on GitHub

(Alternatively, if the embed tag doesn't work out of the box, here is the direct link: GaiaOS GitHub Repository)

How I Built It

I engineered GaiaOS over a high-velocity weekend sprint, keeping both performance and aesthetics at the forefront.

Vite + React Core: Choosing Vite ensured the application is incredibly fast and highly optimized. It allowed for rapid module-reloading during the intense UI development phase.
Immersive Design System: To establish a true enterprise feel, I opted out of standard component libraries and meticulously crafted a custom design system using Vanilla CSS. I implemented heavy use of "Glassmorphism" (complex backdrop blur overlays), distinct space-dark palettes, and modern typography pairs (Inter & Outfit).
Alive UI with Framer Motion: The interface utilizes AnimatePresence to orchestrate fluid, buttery-smooth page transitions that make the application feel truly 'alive.'
Simulated Data Engineering: I utilized Recharts to demonstrate how incoming large-scale data (conceptualized via Snowflake) could be beautifully charted—comparing metrics like Global Carbon Offsets vs. Renewable Energy Adoption.

Highlighted Technical Decisions:

I strongly believe that AI agents must be fundamentally secure. Building the Auth0Gate component was an exciting challenge. Instead of a typical login page, I built an interactive Terminal Simulator that visually replicates an OAuth 2.0 Machine-to-Machine (M2M) handshake. It proves the concept of how a sensitive Planetary Oracle should enforce strict scopes before accepting queries.

Prize Categories

I am proudly submitting GaiaOS for the following prize categories:

Best Use of Google Gemini: The core intelligence loop of GaiaOS is powered directly by the official @google/generative-ai SDK. By connecting to gemini-1.5-flash, the Gemini Oracle component acts as a high-speed planetary analyst, streaming actionable sustainability answers dynamically to the user interface.
Best Use of Auth0 for Agents: I tackled this category by creating an immersive developer integration flow. The app demonstrates the architectural concept of securing AI models by explicitly simulating an M2M JWT negotiation handshake before unlocking access to the Gemini endpoint.
Overall Earth Day Winner: Combining highly topical "green tech" focus with a god-tier presentation.

Built by Sivakumar for the DEV Weekend Challenge: Earth Day Edition. Let's protect our planet! 🌍
*Built with ❤️ for the Planet.

Anime Nexus — a sleek community planner for anime fans

Siva Kukkuluri — Sat, 28 Feb 2026 12:29:25 +0000

Anime Nexus — A sleek discovery hub for anime communities

devchallenge #weekendchallenge #showdev #webdev #react

This is a submission for the DEV Weekend Challenge: Community.

The Community

I built this for the massive, passionate, and constantly growing anime community. Whether fans are debating Bleach: Thousand-Year Blood War theories, ranking Naruto jutsu tier lists, or hyping up the latest Solo Leveling episode, organizing those discussions can be a headache.

Fan wikis are too static, and massive Discord servers often feel like shouting into the void. Important discussions and great show recommendations get buried instantly in fast-moving chat threads.

I wanted to build something specifically for anime fans that prioritized clean design, easy discovery, and focused community hubs.

What I Built

I created Anime Nexus — a UI-focused community directory designed to be the ultimate jumping-off point for anime fans to find their specific tribe.

It includes:
📺 Curated Club Discovery

Browse beautifully designed cards highlighting different fan clubs.
See match percentages and active member counts at a glance.
Filter by popular genres (Action, Shonen, Slice of Life).

💬 Live Global Chat

A persistent, real-time sidebar chat.
Users can stay connected to the global conversation without losing their place while browsing the directory.

✨ Frictionless Recommendations

A built-in, lightweight engine to instantly drop a show recommendation to the community feed.

🎯 Specialized Hubs

Curated sections for specific viewing habits (e.g., Weekly Simulcasts, Manga Readers, Cosplay Hub).
"Create your own club" CTA for finding groups for niche or obscure titles.

The goal was to make a practical, visually appealing, and friction-free way to connect fans with the exact groups they are looking for.

Demo

Live Preview URL: Anime Nexus Live App

Live Preview URL: Anime Nexus Live App

Code

07MADARA / anime-nexus

React + Vite

This template provides a minimal setup to get React working in Vite with HMR and some ESLint rules.

Currently, two official plugins are available:

@vitejs/plugin-react uses Babel (or oxc when used in rolldown-vite) for Fast Refresh
@vitejs/plugin-react-swc uses SWC for Fast Refresh

React Compiler

The React Compiler is not enabled on this template because of its impact on dev & build performances. To add it, see this documentation.

Expanding the ESLint configuration

If you are developing a production application, we recommend using TypeScript with type-aware lint rules enabled. Check out the TS template for information on how to integrate TypeScript and typescript-eslint in your project.

View on GitHub

Why this exists

Finding niche anime watch-groups and dropping quick, highly-visible recommendations shouldn't require scrolling through endless generalized chat channels. Anime Nexus keeps it clean and simple with a modern, single-page application layout.

Features

Hero dashboard with community stats
Filterable club directory grid
Submit form for instant anime recommendations
Simulated real-time global chat sidebar
Specialized community collections
Responsive UI with soft-shadow aesthetics

Tech Stack

React * Vite (for lightning-fast development)
Tailwind CSS v4 (for rapid styling and UI components)
GitHub Pages (for automated deployment)

Run locally

Clone the repository: git clone https://github.com/07MADARA/anime-nexus.git
Navigate to the folder: cd anime-nexus
Install dependencies: npm install
Start the server: npm run dev

Ideas for next version

Backend Integration: Hook up Firebase or Supabase to make the live chat truly multiplayer.
Authentication: Allow users to log in, save their favorite clubs, and create persistent club cards.
Dynamic Search: Implement a live search bar for the club directory.

PantryPulse: Ending Food Waste with Real-Time Community Coordination 🥫🧡

Siva Kukkuluri — Fri, 27 Feb 2026 16:25:01 +0000

This is a submission for the DEV Weekend Challenge: Community

The Community

Mutual aid organizers, local volunteers, and neighbors working together to fight food insecurity. Many cities rely on "Community Fridges" or "Free Little Pantries" to provide barrier-free access to food.

However, there is a massive coordination problem. Donors often drop off perishable food at fridges that are already full, leading to waste. Conversely, people in need often walk miles to a location only to find it completely empty. The community needs a way to broadcast the real-time status of these resources.

What I Built

I built PantryPulse, a mobile-first web app that acts as a real-time health monitor for local mutual aid pantries.

Anyone walking past a community fridge can simply open the app and tap "Empty," "Low," or "Stocked." This crowdsourced data instantly updates the dashboard, allowing donors to route their groceries to the locations that need them most urgently, and saving those in need from wasted trips.

Demo

🔗 Live Prototype: https://07MADARA.github.io/pantry-pulse/
antry-pulse/

(Try clicking the status update buttons to see the dashboard metrics shift in real-time!)

Code

07MADARA / pantry-pulse

How I Built It

Frontend: Designed with a mobile-first approach using pure HTML and Tailwind CSS. I focused heavily on accessibility and clear visual indicators (color-coded statuses and iconography) so the app can be read at a glance while walking outside.
Logic: I used Vanilla JavaScript to handle the real-time DOM manipulation.
Data State: To ensure this MVP was instantly usable and verifiable by judges without standing up a backend database, I implemented browser localStorage. This allows the status updates to persist, proving the core functionality of the crowdsourcing concept.

OppDrop: A Peer-to-Peer Opportunity Exchange for Job Seekers 🤝💼

Siva Kukkuluri — Fri, 27 Feb 2026 15:52:11 +0000

This is a submission for the DEV Weekend Challenge: Community

The Community

For fresh graduates and students—especially those aiming for highly competitive government exams (like GATE) or public sector roles—finding the right opportunity is a scattered mess. You have to constantly check dozens of separate organization portals, employment news sites, and company boards for notifications that often close quickly. I built this for the community of job seekers who rely on peer-to-peer sharing to find off-campus drives, part-time gigs, and hidden roles before deadlines hit.

What I Built

I built OppDrop (Opportunity Drop), a sleek, fast, community-driven job exchange board. Instead of relying on a centralized aggregator, OppDrop allows users to share direct application links to opportunities they find in the wild. I included a quick-filter system so users can instantly sort by "Govt / PSU," "Private Sector," or "Part-Time / Remote" roles to cut through the noise.

Demo

🔗 Live Application: https://07MADARA.github.io/opp-drop/

(Drop a test link and see how the filter system works!)

Code

07MADARA / opp-drop

How I Built It

For this second weekend sprint, I focused on building a clean UI with functional data sorting.

Frontend: I utilized pure HTML and Tailwind CSS to build a modern, card-based interface with distinct color-coding for different job categories.
Logic & Data: Vanilla JavaScript handles the core functionality, including the category filtering system. To keep the app lightweight and instantly deployable without a backend, I used localStorage to persist the job listings and categories on the user's local machine, serving as a rapid prototype database.

CAD Connect: A Community Knowledge Board for Mechanical Designers ⚙️

Siva Kukkuluri — Fri, 27 Feb 2026 14:58:15 +0000

This is a submission for the DEV Weekend Challenge: Community

The Community

While software engineers have an abundance of platforms to share code snippets and debug errors, mechanical engineering and product design students often struggle to find a centralized place for quick knowledge sharing. When working on complex assemblies—like running static and modal analysis on compressor models or trying to map multistage cyclic symmetry in Solid Edge—finding the right material parameter, tolerance standard, or software shortcut usually requires digging through outdated, clunky forums. I wanted to build a clean, modern hub specifically for this community.

What I Built

I built CAD Connect, a lightweight community board designed specifically for CAD enthusiasts and mechanical designers. It allows users to quickly post and categorize design tips, keyboard shortcuts, and analysis parameters. The goal was to eliminate the friction of sharing quick, highly technical mechanical design knowledge with peers.

Demo

🔗 Live Application: https://07MADARA.github.io/cad-connect/

(Feel free to test it out by dropping a post!)

Code

07MADARA / cad-connect

How I Built It

For this weekend sprint, I wanted to focus on speed, simplicity, and a clean UI.

Frontend: I used pure HTML and Tailwind CSS (via CDN) to rapidly prototype a modern, responsive interface.
Logic & Data: I used Vanilla JavaScript to handle the dynamic rendering of the community feed. To ensure the app functioned as a true prototype without spinning up a backend server for the weekend, I utilized the browser's localStorage to manage state, allowing users to create posts and add "likes" that persist across their local sessions.

Predicting Industrial Heartbeats: Building a RAG Pipeline for Mechanical RUL with Elastic

Siva Kukkuluri — Tue, 24 Feb 2026 09:31:37 +0000

Disclaimer: This blog post was submitted to the Elastic Blogathon Contest and is eligible to win a prize.

Predicting Industrial Heartbeats: Building a RAG Pipeline for Mechanical RUL with Elastic

In the world of Mechanical Engineering, an unplanned machine failure is more than just a repair bill—it’s a disruption to the entire production lifecycle. As a final-year student at Pragati Engineering College, I’ve spent months exploring how to turn raw sensor data into actionable intelligence.

The Challenge: From Reactive to Proactive

Traditional maintenance relies on simple thresholds (e.g., "if temperature > 90°C, stop"). However, machinery degradation is often subtle and nonlinear. My project, the Autonomous Mechanical Health & RUL Monitor, seeks to solve this by identifying early degradation patterns using AI.

Why "Vectorized Thinking" with Elastic?

To move beyond simple alerts, we need to treat machine states as high-dimensional data. This is where Elasticsearch shines as a vector database. By converting sensor readings—vibration, temperature, and pressure—into vectors, we can perform Semantic Search on physical hardware states.

Vector Search: Instead of searching for exact numbers, we search for "health patterns" that look like previous failure modes.
Hybrid Search: Combining Elastic’s keyword search (for specific machine IDs) with vector search (for health status) ensures 100% accuracy in retrieval.

The Architecture: RAG for Remaining Useful Life (RUL)

The "brain" of this system is a Retrieval-Augmented Generation (RAG) pipeline. Here is how it works:

Ingestion: Sensor data from my 4-stage axial compressor model is vectorized and stored in Elastic.
Retrieval: When a query is made about a specific component's health, Elastic retrieves the most similar historical degradation paths.
Generation: This context is fed into an LLM to predict the Remaining Useful Life (RUL) in a human-readable report.

Technical Snippet: Connecting the Dots

Below is a simplified look at how we utilize the Elastic vector store within our Python environment:

# Initializing Elastic Vector Store for RUL Data
from elasticsearch import Elasticsearch

es = Elasticsearch("your-cloud-id", api_key="your-api-key")

# Example Vector Query for Machine Health
query_vector = [0.12, 0.45, -0.03, ...] # Vectorized sensor state

response = es.search(
    index="mechanical-health",
    knn={
        "field": "sensor_vector",
        "query_vector": query_vector,
        "k": 5,
        "num_candidates": 100
    }
)
print("Similar failure modes found:", response['hits']['hits'])

Conclusion

By leveraging Elasticsearch’s vector capabilities, we can bridge the gap between heavy machinery and advanced AI. For the future of Industry 4.0, every vector tells a story of machine health. I'm excited to keep pushing the limits of what Mechanical Engineers can build with the Elastic Stack.

Published for the Elastic Blogathon 2026. #ElasticSearch #VectorSearch #RAG #MechanicalEngineering #Industry40