DEV Community: Aditya Nakhale

EventGenie: Gen AI-Powered AI Event Planner

Aditya Nakhale — Thu, 29 Oct 2026 11:05:10 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 4: Ghost in the Machine

What I Built

I built EventGenie — an intelligent, full-stack AI celebration and event planner designed to take the friction, guesswork, and financial stress out of organizing milestone events. Whether it's a dream wedding, a milestone birthday party, or a vibrant community gathering, EventGenie transforms raw ideas into actionable, comprehensive plans in under 30 seconds.

Key Features:

🧠 4-Step AI Event Wizard: Guides users through event type, guest count, location, priority preferences, and aesthetic vibes.
💰 Smart Budget Allocator: Uses Google Gemini to intelligently break down total budgets into granular vendor categories, complete with interactive, real-time rebalancing sliders.
📋 Dynamic Vendor Checklist: Auto-generates categorized vendor recommendations with itemized cost estimates, priority badges, and booking status tracking (Not Started, Contacted, Booked, Paid).
📅 Milestone Timeline & Countdown: Generates an interactive month-by-month and week-by-week schedule leading up to the celebration day, with automated overdue detection.
💡 Personalized Creative Ideas: Curates 4 distinct idea buckets — Themes, Décor, Catering/Menu, and Entertainment.
💬 Context-Aware AI Chat Concierge: A real-time assistant that holds full context of the active event (budget, guest count, venue, style) to answer custom questions and give cost-saving tips.
💾 Local-First Privacy: Complete plan persistence with zero server-stored PII, browser-isolated localStorage, and instant JSON export/import.

Demo

Live Application: https://ais-pre-75d6by5g3id7vwegimrcm6-679570267874.asia-southeast1.run.app
Development App: https://ais-dev-75d6by5g3id7vwegimrcm6-679570267874.asia-southeast1.run.app

Code

GitHub Repository: https://github.com/AdityaNakhale/Gen-AI-event-planner-.git

The codebase is built with clean Vanilla ES Modules, Vite, Node.js Express (server.ts), and 133 unit tests (100% pass rate) with Jest.

How I Built It

EventGenie combines modern web engineering with state-of-the-art GenAI integration:

AI Framework & Model Routing:
- Built around the official TypeScript SDK on a secure Node.js Express proxy layer (server.ts).
- Implemented dynamic multi-model redundancy using gemini-2.5-flash as primary, falling back to gemini-flash-latest and gemini-3.8-flash for ultra-fast, zero-downtime inference.
Defensive AI Engineering & Deterministic Fallbacks:
- Structured JSON schema enforcement with regex boundary bracket extraction (search(/[{[]/)) and markdown code-fence sanitization.
- Built with a zero-crash guarantee: If API limits or network dropouts occur, deterministic mathematical algorithms immediately step in to allocate proportional category splits and timeline milestones without breaking the user experience.
Enterprise Security & Perimeter Defense:
- Rate limiting (60 requests/minute per IP) to mitigate denial-of-service attempts.
- Zero-trust key isolation: GEMINI_API_KEY is strictly confined to server-side environment variables and never exposed to client bundles.
- Strict Content Security Policy (CSP), X-Content-Type-Options: nosniff, and zero server PII storage.
Testing & Quality Assurance:
- 133 Jest unit tests covering budget math integrity, timeline date calculations, and localStorage CRUD operations.

Why Does Open Innovation Matter?

Open innovation bridges the gap between high-level generative AI capabilities and everyday human needs. When models and open frameworks allow flexible, structured execution:

Democratizing Professional Event Planning: Professional event planners charge thousands of dollars, making structured planning inaccessible to families on tight budgets or grassroots community organizers. Open AI tooling allows us to package high-level organizational intelligence into a free, accessible web app.
Privacy and Data Sovereignty: By coupling open architecture with local-first client storage, users retain 100% ownership of their celebration plans without handing sensitive personal schedules, guest lists, or financial data to opaque, ad-driven databases.
Resilience and Modularity: An open design allowed us to decouple UI, business math, and LLM inference, ensuring the planner remains functional, responsive, and rebalanceable even without an active internet connection.

My Agent Session

Built and iterated with the assistance of Google AI Studio Build Agent, orchestrating the transition from initial scaffolding to a hardened full-stack architecture with rate limiting, comprehensive Jest test coverage, and WCAG AA accessibility compliance.

The Spirit of It

Planning a celebration is supposed to be joyous, but in reality, it's often overwhelmed by spreadsheets, chaotic vendor notes, and budget anxiety.

The moment EventGenie landed was when we tested it with a complex, real-world scenario: "A beach wedding with 120 guests and a tight budget that needs vegetarian catering and live acoustic music." In under 15 seconds, the AI not only distributed every single rupee/dollar with zero rounding drift, but it also suggested hiring a local food truck collective to preserve the food budget for photography, and generated a 12-month countdown timeline that told the couple exactly what to do first. It replaced weeks of decision fatigue with instant clarity and excitement.

Prize Categories

Best Overall Application
Most Creative Use of Gemini API
Best UX & Accessibility

Author / Team:

Aditya Nakhale (adityanakhale0@gmail.com)

Food Rescue AI - Hacktoberfest Submission

Aditya Nakhale — Wed, 14 Oct 2026 16:26:33 +0000

Liquid syntax error: Tag '{% https://github.com/AdityaNakhale/FoodRescuAI-SpringBoot.git %}' was not properly terminated with regexp: /\%\}/

FoodRescue AI: Mobilizing Real-World Volunteers with Local Open-Weight Routing Agents

Aditya Nakhale — Tue, 06 Oct 2026 11:22:03 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

What I Built

FoodRescue AI is an automated rescue and dispatch platform designed to prevent edible food from ending up in landfills while mobilizing volunteers into the physical world.

Rather than keeping community members glued to screens or browsing endless feeds, FoodRescue AI's objective is to make the digital experience under 60 seconds so people can head outside, navigate their neighborhoods, and deliver food to shelters. When commercial kitchens, restaurants, or event venues log surplus food, the system evaluates urgency, batch-allocates geographically aligned pickup points, and delivers precise dispatch itineraries straight to active local couriers and volunteers.

Demo

Repository Walkthrough: Check out the architecture, schema setup, and distance calculation services directly on GitHub.
Demo Video / Live Link: (Add your hosted demo URL or screen recording link here)

Code

The project is open source and hosted on GitHub:

AdityaNakhale / FoodRescuAI-SpringBoot

#FoodRescuAI-SpringBoot ##Food Rescue AI is a smart web-based platform that connects food donors with nearby NGOs to reduce food wastage It uses location-based intelligence (latitude & longitude) to allocate surplus food efficiently. Donors can post food details while NGOs receive optimized assignments based on distance and capacity. The system is built using Spring Boot, MySQL, and Thymeleaf It ensures fast, fair, and real-time food distribution. The goal is to create an efficient food rescue network and support communities in need.

#Technologies Used -Frontend: HTML, CSS, JavaScript, Thymeleaf -Backend: Java, Spring Boot -Database: MySQL -Tools & APIs: Google Maps API (for location handling), REST APIs

#Key Features 🔐 Secure login system for Donors and NGOs 📍 GPS-based location tracking (Latitude & Longitude) 🤖 Intelligent food allocation algorithm 📦 Real-time food donation posting and management 📊 Dashboard for tracking donations and requests 📁 File upload support (for NGO verification) ⚡ Fast backend…

View on GitHub

How I Built It

The architecture focuses on fast local processing, predictable allocation routing, and an accessible mobile UI:

AI Urgency & Triage Logic: Analyzes incoming donation parameters (spoilage windows, storage conditions, packaging type) using open-weight language models via Ollama to determine triage priorities and create concise pickup instructions.
Backend Service: Engineered with Java 21 and Spring Boot, featuring transactional persistence via Spring Data JPA and relational storage powered by MySQL.
Geographic Distance Routing: Integrates geographical distance evaluation to cluster nearby food pickup locations, matching them with local shelter drop-offs and volunteer operational radii to minimize transit delays.
Volunteer Web UI: Designed with Thymeleaf and Tailwind CSS, built mobile-first with high daylight readability so volunteers can easily read dispatch cards while walking or cycling outdoors.

Why Does Open Innovation Matter?

Zero Ongoing Cost for Non-Profits: Grassroots shelters and food rescue programs run on constrained budgets. Relying on closed, pay-per-token proprietary APIs creates continuous overhead. Open-weight models and self-hosted tools run at virtually zero operational inference cost.
Privacy for Vulnerable Communities: Food rescue logistics handle residential pickup addresses, shelter locations, and recipient identities. Running self-hosted open-source software ensures that community data never leaves trusted local environments.
Edge Resilience: Community aid must continue even during internet dead zones or regional server disruptions. Open-source stacks can be deployed locally on laptops or offline edge servers without external cloud dependencies.

Prize Categories

Best Use of Gemma (Structured triage and urgency ranking)
Best Use of Render (Cloud deployment and backend hosting)