Kavach-AI

Our AI guardian watches, warns, and prevents scams automatically.

Comment

Inspiration

People receive messages, notices, and information every day, but understanding what matters and deciding what action to take is often difficult. We wanted to build something that helps people understand information instead of feeling overwhelmed by it.

One of our team members wants to become a Cybersecurity Manager in the future, which inspired us to think more deeply about trust, risk awareness, and helping users make informed decisions. That idea became KAVACH AI.

What it does

KAVACH AI is an AI-powered decision support platform that analyzes user-provided messages and transforms them into understandable guidance.

Users receive:

  • Context understanding
  • Risk awareness
  • Clear explanations
  • Recommended next steps

Instead of replacing decisions, KAVACH AI helps users understand information while keeping humans in control.

How we built it

KAVACH AI was developed as a browser-based prototype using HTML, CSS, and JavaScript with Gemini API integration.

The workflow follows:

User Input → AI Understanding → Analysis → Guidance → Human Decision

For this prototype, users enter their own API access during testing instead of storing credentials through backend infrastructure.

Challenges we ran into

One challenge was designing an AI system that provides useful guidance without encouraging overdependence on automation.

We also faced technical constraints while building. Initially, we planned backend integration, but due to limited hackathon time we focused on creating a lightweight browser-based prototype and implemented the project mainly using JavaScript.

Accomplishments that we're proud of

We are proud that KAVACH AI became a working prototype rather than remaining an idea.

We also built the project around explainability, human oversight, and responsible AI principles instead of focusing only on technical complexity.

What we learned

This project taught us that building AI is not only about generating outputs but also about building trust and creating useful user experiences.

We learned how AI systems, interface design, and responsible decision-making work together to create meaningful products.

What's next for KAVACH AI

Future improvements include secure backend infrastructure, stronger authentication, and improved architecture.

We also plan to add multilingual support, accessibility improvements, and more real-world use cases to expand KAVACH AI's impact.

Built With

Share this project:

Updates