SCAMSHIELD
ScamShield combines a stunning, high-performance UI with a robust, heuristically-driven AI backend. π¨ Premium Frontend (Client) Deep-Space Dark Aesthetic: Custom UI featuring Abyss Black (#050507) backgrounds with vivid Signal Green (#00d992) neon highlights. Glassmorphism & Motion: Immersive component design featuring frosted glass effects, glowing hover states, and smooth 60fps micro-animations. Dynamic Threat Workflow: Interactive SVG pipelines and infinite marquee scrolling displaying real-time system capabilities and supported platforms. π§ Intelligent Backend (Server) π URL Analysis: Detects IP-based URLs, shorteners, suspicious TLDs, and phishing keywords. Integrates with Google Safe Browsing. π¬ Message Analysis: Deep keyword matching, urgency detection, and OpenAI intent classification. π± QR Code Scan: Decodes QR codes from uploaded images to analyze hidden URLs. πΌοΈ Image Analysis: AI-powered scam content detection using OpenAI Vision. π Heuristic Scoring: Returns a human-readable Risk Score (0β100) for every analyzed vector.
Om PandeyAi_Portfolio_
An AI-powered personal portfolio built with Google Gemini API to create an interactive and personalized experience. The portfolio showcases the developerβs skills, projects, achievements, and experience, while Gemini enables intelligent interactions and dynamic responses based on the userβs queries. The project focuses on combining modern UI/UX, AI interaction, and real-world usability into a single professional portfolio experience.
Supreet SinhaAgentTown
AgentTown is a multi-agent AI platform where specialized AI agents collaborate, communicate, debate, and resolve conflicts to solve complex real-world problems. It uses an Orchestrator to delegate tasks to Research, Finance, Planning, Marketing, Operations, and Risk agents. The agents analyze problems from different perspectives, adapt to changing constraints through What-If scenarios, and combine their insights into one practical and actionable solution.
NEERAJ YADAVElectroGuard AI
ElectroGuard AI doesn't just react to electrical faultsβit predicts them. By combining live AR safety boundaries for citizens, multimodal AI triage for repair crews, and predictive weather modeling for power departments, ElectroGuard AI turns every smartphone into an active guard for the electrical grid.
Mahi PanwarCertifyPro
CertifyPro is a smart and easy-to-use certificate generation and management platform that helps organizations create, personalize, and send certificates efficiently. Users can generate certificates individually or in bulk, automatically personalize recipient details, and send certificates directly via emailβsaving time and reducing manual work.
Pritam KumarPitchPanel-Ai
Inspiration & Problem: Over ninety percent of startups fail simply because founders build products that nobody actually wants or is willing to pay for. Early-stage validation is fundamentally broken: friends and family offer polite encouragement that creates a false sense of security, while experienced investors and domain experts are inaccessible to first-time founders. We created PitchPanel AI to bridge this gap by simulating a diverse advisory board capable of stress-testing any startup or product concept with raw, multi-angle scrutiny in seconds. What It Does & How It Works: When a user submits an idea on PitchPanel AI, the platform evaluates it from three opposing viewpoints simultaneously: an Investor analyzing market size, unit economics, and funding viability; a Skeptical Customer reacting with candid, price-conscious hesitation; and a Competitor detailing existing solutions and market moat vulnerabilities. Once all three perspectives are gathered, an AI Moderator synthesizes the feedback into a final verdict featuring a viability rating out of ten, the single biggest strength, the primary risk, and an immediate next validation step. The application features user authentication, chat history preservation, and rolling twenty-four-hour rate limits to protect API resources. Technical Implementation & Architecture: PitchPanel AI is built with Python and Django 5 on the backend, handling user sessions, database integrity, and RESTful API endpoints. To ensure fast response times, the backend employs Python's ThreadPoolExecutor to dispatch the three persona prompts concurrently to Google Gemini's REST API, cutting total waiting time by over sixty percent before triggering the final verdict. Data persistence is engineered with dual-database portability, allowing zero-setup local testing on SQLite while scaling to Supabase PostgreSQL in production. The frontend features a sleek dark glassmorphism aesthetic built with Bootstrap 5, responsive off-canvas sidebars, and asynchronous JavaScript fetch calls for a fluid, real-time chat experience without page reloads. Challenges, Accomplishments & Future Roadmap: During development, we navigated high-throughput concurrency by upgrading to Google's gemini-flash-lite-latest model and building an automated exponential backoff system with model fallbacks to eliminate rate-limit errors. We also resolved serverless database connectivity by routing traffic through Supabase's transaction connection pooler to bypass IPv6 limitations on cloud platforms like Vercel. We are proud to have built a resilient, production-ready full-stack application that produces truly divergent, non-generic AI feedback. Moving forward, we plan to introduce one-click exportable PDF investor summaries, live competitive pricing scrapers, and customizable industry-specific advisory personas.
Som ThakurSuchakAI
SuchakAI is an AI-powered personalized government-scheme discovery and guidance platform designed to bridge the gap between government benefits and the citizens who are eligible to receive them. India has numerous government schemes, scholarships, subsidies, grants, employment programs, and welfare initiatives, but many citizens struggle to discover the schemes relevant to them, understand complex eligibility criteria, or complete the application process on time. As a result, available benefits can remain underutilized. SuchakAI solves this last-mile information gap by creating a personalized dashboard for every user. After providing basic information such as state, age, occupation, education, income range, and areas of interest, users automatically see the government schemes most relevant to their profile whenever they open the platform. The system uses **deterministic eligibility and filtering logic** to identify and rank relevant schemes, while the **Gemini API** is used for intelligent tasks such as understanding complex government documents, interpreting eligibility conditions, and generating simple explanations of why a particular scheme may be relevant to the user. For each scheme, SuchakAI provides the key information a citizen needs: benefits, eligibility criteria, required documents, application steps, deadline or status, and the official application/source link. The platform also emphasizes trust by linking information to official sources and tracking when scheme information was last verified. The core idea behind SuchakAI is simple: **Government benefits should not depend on citizens knowing where and how to search for them.** SuchakAI transforms fragmented and complex public information into **personalized, understandable, and actionable opportunities**, helping citizens discover what they may be entitled to and understand what they should do next.
Priyanshu RajHackord
Hackord is an AI-powered collaboration and project-management platform built specifically for hackathon participants and student developers. It brings the complete hackathon workflow into a single workspaceβfrom discovering hackathons and finding teammates to building projects, collaborating, managing deadlines, preparing presentations, and staying updated. Key Features Hackathon Discovery: Explore upcoming hackathons with event details, deadlines, prizes, eligibility, and registration links. Student & Team Discovery: Search registered students by name, skills, college, and experience, view developer profiles, and connect with potential teammates. Hackord Rooms: Create dedicated project workspaces with team members, tasks, milestones, timelines, links, GitHub repositories, and project-specific collaboration. Real-Time Collaboration: Support team communication through general chat, direct messaging, and audio/video collaboration. GitHub Integration: Connect project repositories and developer profiles to keep coding activity and project information accessible within the workspace. Gemini AI Assistant: Get AI-powered help for ideation, technical queries, project planning, task breakdown, summaries, architecture, and hackathon preparation. AI PPT Generation: Generate structured pitch-deck content from project information, including slide structure, key points, speaker notes, and presentation-ready content. Smart Reminders & Automation: Automate deadline and milestone notifications through Email and WhatsApp workflows. Centralized Dashboard: Track upcoming hackathons, active rooms, tasks, deadlines, notifications, and recent project activity from one place. Admin & Data Management: Manage users, rooms, hackathon information, submissions, and platform activity through an administrative dashboard. Technology Stack & Services Frontend: React 19, Vite, TypeScript, Tailwind CSS, TanStack Router/Query, Radix UI, Three.js Backend: Node.js & REST APIs Database: MongoDB AI: Google Gemini API Communication: Agora RTC for audio/video collaboration Automation: ViaSocket for Email, WhatsApp, and Google Sheets workflows Developer Integration: GitHub Presentation Generation: PptxGenJS Deployment: Vercel Hackord is designed to reduce the need to switch between multiple tools such as Discord, WhatsApp, project-management tools, GitHub, AI assistants, presentation tools, and calendar/reminder services by bringing the essential hackathon workflow into one unified platform. The current project repository documents Hackord as a collaboration workspace for discovering hackathons, forming teams, creating project Rooms, tracking deadlines, discovering builders, managing notifications, and collaborating on projects, with the listed React/Vite/TanStack/Tailwind/Three.js ecosystem.
Safal TiwariScamShield AI
ScamShield AI is a full-stack web platform that helps job seekers instantly detect fraudulent job and internship offers. Users paste in a job posting's text or upload a screenshot/PDF of an offer letter, and the app analyzes it for known scam patterns β upfront payment demands, informal Telegram/WhatsApp-only recruitment, instant hiring with no interview, "earn daily" task scams, and spoofed company domains. Built with a React and Tailwind CSS frontend, a Node.js/Express backend, MongoDB for storage, and Google's Gemini API for multimodal analysis, it returns a clear risk score, verdict, and specific red/green flags with evidence, plus practical steps to verify the offer β giving users a fast, evidence-based check before they fall victim to a scam.
Harsh ShrivastavaSevaSetu
SevaSetu is an AI-powered platform for handling citizen complaints. A citizen can submit a complaint through text along with image and their location. Gemini AI understands the complaint, identifies the problem, and determines its priority. The system then routes the complaint to the appropriate department. If multiple people report the same issue from a nearby location, spatial clustering helps identify them as one common issue. Finally, the concerned officer verifies and resolves the complaint, while the citizen can track its status.
Roshan KumarFarmDirect
# πΎ FarmDirect ### From Farm to Direct β Connecting Farmers and Consumers Through Technology FarmDirect is a **full-stack, AI-powered agricultural marketplace** that aims to connect farmers directly with consumers and reduce the gap created by traditional agricultural supply chains. In a conventional supply chain, agricultural products can pass through several intermediaries before reaching the consumer. This can make it difficult for farmers to access a wider market and can increase the final price paid by consumers. FarmDirect provides a digital platform where farmers can showcase their produce directly and consumers can discover and purchase agricultural products through a unified marketplace. --- ## π― The Problem Farmers play a critical role in our food system, but many face challenges such as: * Limited access to wider markets * Dependence on intermediaries * Difficulty reaching consumers directly * Limited visibility into customer demand * Lack of simple digital tools for managing their produce Consumers face a different side of the problem: * Difficulty discovering local farmers * Limited information about the source of products * Multiple layers between producers and consumers * Less transparency in the agricultural supply chain FarmDirect addresses this disconnect by creating a **digital bridge between producers and consumers**. --- ## π‘ Our Solution FarmDirect provides two primary experiences: ### π¨βπΎ For Farmers Farmers can: * Create and manage their profile * Add agricultural products * Specify product quantity and pricing * Manage their listed produce * Make their products discoverable to consumers * Manage incoming orders ### π For Consumers Consumers can: * Browse agricultural products * Search and discover produce * View product information * Learn about the farmer behind the product * Select products and quantities * Place orders through the platform This creates a simple digital flow: **Farmer β FarmDirect Marketplace β Consumer** instead of relying entirely on a long traditional chain. --- # π€ AI Integration One of the key features of FarmDirect is its **AI-powered Farm Assistant**, powered by the Gemini API. Rather than adding AI only as a demonstration feature, we integrated it as an interactive assistant that can help users access agricultural information through natural-language interaction. Users can communicate with the assistant conversationally instead of having to navigate through complicated information. This creates an opportunity to make agricultural technology more accessible, especially for users who may not be comfortable with complex digital interfaces. --- # βοΈ How It Works At a high level, FarmDirect follows a full-stack architecture: ```text βββββββββββββββββββ β User β ββββββββββ¬βββββββββ β βΌ βββββββββββββββββββ β Frontend β β Web Interface β ββββββββββ¬βββββββββ β API Requests β βΌ βββββββββββββββββββ β Backend β β Business Logic β β Authentication β β REST APIs β ββββββββββ¬βββββββββ β βΌ βββββββββββββββββββ β Database β β Users / Productsβ β Orders / Data β βββββββββββββββββββ + β βΌ βββββββββββββββββββ β Gemini AI API β β Farm Assistant β βββββββββββββββββββ ``` The frontend communicates with the backend through APIs. The backend handles authentication, business logic, product operations, and order-related functionality, while persistent application data is maintained in the database. The AI assistant communicates with the Gemini API to provide conversational assistance. --- # β¨ What Makes FarmDirect Different? FarmDirect is not simply an e-commerce website with agricultural products. Our focus is on combining: ### π± Direct Farmer Access Giving farmers a dedicated digital channel to reach potential customers. ### π Transparency Connecting product information with information about its producer. ### π€ AI Assistance Using conversational AI to make agricultural information easier to access. ### π₯ Two-Sided Marketplace Designing separate workflows for both farmers and consumers. ### π Future Data-Driven Agriculture The platform can evolve beyond buying and selling into a system that uses marketplace data for demand insights, recommendations, and forecasting. Our goal is to use technology to address a real-world agricultural problem rather than building a technology demo without a practical use case. --- # π οΈ Technology Stack FarmDirect is built using modern web technologies. ### Frontend * React * JavaScript * Modern responsive UI * Component-based architecture ### Backend * Node.js * Express.js * REST APIs * Authentication and server-side business logic ### AI * Google Gemini API * AI-powered conversational Farm Assistant ### Database & Data Management * Persistent user, product, and order data * Structured backend data management ### Development & Deployment * Git & GitHub for version control * Vercel for deployment --- # π Authentication & Security FarmDirect includes authentication and protected application functionality to ensure that users can access features appropriate to their role. The backend is responsible for validating requests and managing protected operations instead of relying solely on frontend restrictions. Sensitive credentials such as API keys are kept outside the source code using environment variables. --- # π§βπ» User Journey ### Farmer Journey ```text Register / Login β Create Profile β Add Produce β Set Price & Quantity β Publish Product β Receive Orders ``` ### Consumer Journey ```text Register / Login β Browse Marketplace β Search / Explore Products β View Product & Farmer Details β Select Quantity β Place Order ``` ### AI Journey ```text User Question β Farm Assistant β Gemini API β AI-Generated Response β User ``` --- # π Real-World Impact FarmDirect has the potential to create impact at multiple levels. ### For Farmers * Greater digital market access * More direct customer interaction * Better visibility for their produce * Potential access to future demand insights ### For Consumers * Easier agricultural product discovery * Greater visibility into product sources * Direct interaction with the agricultural marketplace ### For the Agricultural Ecosystem * Encourages digital adoption * Creates opportunities for data-driven decision-making * Provides a foundation for smarter agricultural commerce --- # π Future Roadmap FarmDirect is designed as a foundation that can be expanded significantly. ### Phase 1 β Marketplace β Farmer onboarding β Product listing β Consumer discovery β Order management ### Phase 2 β Connected Commerce * Digital payment integration * Delivery and logistics integration * Location-based discovery * Farmer verification * Ratings and reviews ### Phase 3 β Intelligent Agriculture * AI-powered recommendations * Demand forecasting * Price insights * Farmer analytics * Personalized agricultural assistance ### Phase 4 β B2B Agricultural Network FarmDirect could eventually connect farmers with: * Restaurants * Grocery businesses * Retailers * Hotels * Food processing businesses This could transform FarmDirect from a consumer marketplace into a broader **digital agricultural supply network**. --- # π Why We Built It We built FarmDirect because agriculture is not only a technology opportunity β it is a real-world problem space affecting farmers, consumers, and the broader economy. Instead of building another generic marketplace, we wanted to explore how **full-stack development + AI** could be applied to a meaningful problem. The central idea is simple: > **The person who grows our food should have a more direct path to the person who buys it.** FarmDirect uses technology to shorten that distance. ## π± FarmDirect **From Farm to Direct.**
OM DWIVEDIKUTUMB
KUTUMB is a Family Knowledge & Responsibility Map that helps families understand, connect, and manage important information scattered across documents such as insurance policies, loan statements, utility bills, vehicle documents, and property records. Using Google Gemini's multimodal document understanding, KUTUMB extracts people, entities, relationships, important dates, financial information, and responsibilities from uploaded documents. It then connects information across multiple documents to create a family-level knowledge model with an actionable dashboard, responsibility timeline, recurring obligation predictions, and grounded AI-powered Q&A through Ask KUTUMB. Unlike traditional document storage or reminder apps, KUTUMB focuses on the context behind family information β who owns what, who is responsible, what needs attention, and when action is expected. The goal is to make family knowledge transferable, so important responsibilities don't remain concentrated in one person's head.
Ayush rajPolicy-plus
# Policy+ **Policy+ is an AI-powered policy stress-testing and decision-support platform that helps policymakers explore the potential consequences of policy decisions before implementing them.** Instead of simply showing historical data, Policy+ allows users to define a policy intervention, simulate its measurable impact, compare alternative scenarios, and deliberately stress-test the policy under adverse conditions. The platform uses a **deterministic simulation and risk engine** to calculate metrics such as capacity, utilization, ridership, revenue, operating costs, surplus, and policy risk. It then uses **Google Gemini as an AI Policy Analyst** to interpret these validated results, explain trade-offs, identify important assumptions, and provide understandable insights. The primary demonstration module is a **Bus Policy Stress Tester**, where users can evaluate decisions such as increasing the public bus fleet and observe how the policy behaves under normal, high-demand, high-cost, and combined adverse scenarios. ### Core Features * π **Bus Policy Simulation** β evaluate the impact of fleet and service changes. * π **Scenario Comparison** β compare multiple policy alternatives. * β‘ **Attack My Policy** β deliberately test policies against adverse conditions. * π― **Policy Risk Score** β deterministic 0β100 risk assessment with transparent risk drivers. * π **Sensitivity Analysis** β understand how outcomes change as assumptions vary. * β¦ **Gemini AI Policy Analyst** β explain results, risks, assumptions, and trade-offs using validated simulation outputs. * π§ **Human-in-the-Loop Decision Making** β AI assists analysis but does not make policy decisions. ### Architecture ```text Policy Input β Deterministic Simulation Engine β Scenario & Stress Testing β Policy Risk Engine β Validated Results β Gemini AI Policy Analyst β Insights β’ Risks β’ Trade-offs β Human Decision ``` ### Technology Stack **Frontend:** React, Vite, Tailwind CSS, Recharts, Lucide React **Backend:** Python, FastAPI, Pydantic **AI:** Google Gemini API **Analysis:** Deterministic simulation and rule-based risk modeling ### Key Principle > **Gemini does not calculate the policy metrics. Policy+ calculates the numbers deterministically; Gemini explains what those numbers mean.** Policy+ is designed as a **decision-support prototype**, not a guaranteed forecasting system. Its conclusions depend on the assumptions and scenarios provided by the user. **Tagline:** **Simulate. Stress-Test. Decide.**
Madhav RajFarmer-Consumer-Bond
AgriFresh connects farmers directly with consumers through an AI-powered 3D cold chain digital twin, real-time IoT freshness telemetry, and Web3 escrow trade contracts. 3D Spatial Digital Twin: WebGL/Three.js-powered visualizer for off-grid solar micro-cold hub monitoring and chamber diagnostics. Gemini Multimodal AI: Vision-based crop quality grading and multilingual voice negotiation to bridge rural dialect barriers. IoT Cold-Chain Telemetry: Live tracking of temperature, relative humidity, ethylene levels, and solar battery reserves. Direct-to-Consumer Fair Trade: Eliminates exploitative intermediaries, backed by smart escrow verification for automated payouts upon gate-in quality approval.
Priyanshu SarjanDaily-habit-Remainder-Tracker
Daily Habit Reminder is a web-based application that helps users create, track, and complete their daily habits. The application sends reminders according to the scheduled time and allows users to mark habits as completed using a checkbox.
Rahul DhakarCampus OS
Campus OS is an AI-powered student risk intelligence platform that turns messy academic documents into a personalized risk briefing. It uses Gemini to extract schedules, deadlines and attendance, investigate college policies, identify risks, and simulate the consequences of student decisions.
Divyanshu KubdeSEO_AGENT
SEO Agent is an AI-powered SEO consultant that audits any URL for technical, on-page, and content issues, researches keywords by search intent, and benchmarks the page against top competitors β returning a prioritized action plan instead of generic advice. Built with Node.js, Express, and the Gemini API, with a rule-based fallback engine ensuring it never fails silently.
Devansh PatelCodeArena
CodeArena is a comprehensive full-stack application designed to host competitive programming contests while completely automating tedious backend operations. It eliminates the manual headache of tracking test cases, managing leaderboards, and drafting individual participant certificates by transforming the entire pipeline into a 100% automated, single-click system.
Maruti Guptamental health tracker
chatbot tracker mood
Paridhi Likhitkaryappers
Yappers is an AI-powered learning companion designed to help students study smarter and stay accountable. It combines personalized learning plans, AI communication, Study Mode, and real-time focus monitoring. Students can interact through text or natural voice conversations with concise, supportive guidance. Its camera-based monitoring detects focus, phone usage, and inactivity during study sessions. Yappers transforms learning into a structured, interactive, and personalized experience.
Aman Yatan SenAI-MCQ-PREPRATION-PROJECT
AI-MCQ-PREPRATION THIS PROJECT SPECIALLY FOR STUDENTS FOR PREPRATION MCQ
yogank chaurasiacyberguard ai
its is help to all fraud and online fraud
vedansh KasarAgrimarket
To solve e-commerce of farmer.
Abhay Kumar SinghAI fake information research agent
The agent researches the claim, checks credible sources, compares evidence, and responds with something like: Verdict: Unverified The claim could not be confirmed by reliable sources. More evidence is required before treating it as factual.
Yujal Ahirwar