🌱 Kissa-E-Paisa (The Money Tale)

Building the "Atmanirbhar Investor" through Generative AI

"80% of new demat accounts are from non-metro cities. They don't need another dashboard; they need a guide."

🚀 Click Here to Play the Live Simulation

(No Login Required • Works on Mobile • Zero Cost)

---

🎯 The Problem Statement

India's securities market is exploding with 337 Million households entering the fold. However, new investors—especially from Tier 2/3 cities—face a critical "Behavioral Gap":

1. Fear & Mistrust: Complex jargon makes them feel inadequate.
2. Reckless Speculation: Without guidance, they treat trading like gambling.
3. Language Barrier: Most financial tools are English-first, alienating "Bharat."

💡 The Solution: Kissa-E-Paisa

Kissa-E-Paisa is an AI-powered text-RPG (Role Playing Game) that sits as a layer between the user and the real market.

Instead of boring tutorials, we drop users into realistic financial dilemmas (e.g., “Your cousin asks for a loan” or “Market crashes 10% today”). They make choices, face consequences, and build "Muscle Memory" for risk—all in their local Hinglish dialect.

---

🏆 Why This Wins (Alignment with Ideathon Pillars)

CDSL Pillar	Our Feature	Impact
Inclusion	Hinglish AI Engine	Breaks the language barrier. The AI speaks like a local friend ("Bhai, risk hai!"), not a banker.
Behavioral Insight	Loss Aversion Simulator	We simulate financial loss in the game so users learn to handle panic before risking real capital.
Innovation	Infinite Scenarios	Powered by Google Gemini, no two games are alike. The content is dynamic, not a static quiz.
Trust	Regulatory Guardrails	System prompts strictly prevent specific stock tips, ensuring 100% compliance with SEBI guidelines.

---

⚙️ Technical Architecture & Innovation

Kissa-E-Paisa is built on a Serverless, Event-Driven Architecture designed for high scalability and zero maintenance costs. Unlike traditional rule-based chatbots, we utilize a State-Aware Generative Engine.

🏗️ System Data Flow

graph TD
    subgraph UserLayer ["User Interaction Layer"]
        User([User Mobile/Web])
        Input[/"Vernacular Input (Hinglish)"/]
    end

    subgraph AppLayer ["Application Orchestrator (Streamlit)"]
        Session["Session State Manager"]
        Safety["🛡️ SEBI Compliance Guardrail"]
        PromptEng["Dynamic Prompt Injection"]
    end

    subgraph Intelligence ["Intelligence Layer (Google Cloud)"]
        Gemini[/"Google Gemini 1.5 (LLM)"/]
    end

    User --> Input
    Input --> Session
    Session -->|Context + Wallet Data| PromptEng
    PromptEng -->|Sanitized Request| Safety
    Safety -->|Authorized Prompt| Gemini
    Gemini -->|Financial Simulation| Session
    Session -->|Structured Response| User

Loading