SentryX

• 

  this the front page of our app

In today’s digital age, misinformation and deepfakes are spreading faster than ever, impacting elections, finance, health, and everyday social media interactions. We wanted to create a tool that empowers users to instantly identify false information and manipulated media, giving them a reliable “truth-check” in real-time.

What it does

SentryX is an intelligent AI agent that:

Scans news articles, tweets, and videos.

Detects fake news, biased claims, and AI-manipulated media.

Cross-checks facts using trusted sources like Google Fact Check, Wikipedia, and news APIs.

Provides a real-time truth report with a credibility score, flagged claims, and verified sources.

It’s fully autonomous, easy to use, and acts as a digital sentinel protecting users from misinformation.

How we built it

Maestro Platform: Deployed multiple agents for input, content extraction, deepfake detection, fact-checking, and summarization.

External APIs: Integrated Twitter/X API, YouTube API, Google Fact Check API, and Wikipedia API.

Python & Libraries: Used BeautifulSoup, Newspaper3k, and Maestro LLM for content extraction and summarization.

Workflow: Each agent communicates seamlessly on Maestro, producing a clean, real-time output for any submitted link or media.

Challenges we ran into

Time Constraint: Building multiple autonomous agents in a single day required prioritizing features.

API Limitations: Rate limits on external APIs sometimes slowed down testing.

Deepfake Detection: Integrating reliable AI detection models for videos was challenging within the hackathon timeframe.

We solved these by focusing on a strong MVP, prioritizing real-time analysis for articles and images first, with optional video deepfake detection.

Accomplishments that we're proud of

Built a fully functional, multi-agent system on Maestro within 18 hours.

Created an intuitive workflow that outputs a credibility score and a detailed truth report.

Successfully demoed detection on fake news articles and manipulated media, showing real-world impact.

Delivered a polished GitHub repository with setup instructions, API calls, and architecture diagrams.

What we learned

How to rapidly deploy multiple AI agents on Maestro.

The importance of clear, autonomous agent workflows for real-time applications.

Practical experience integrating external APIs for fact-checking and deepfake detection.

How to prioritize features under time constraints while maintaining functionality and impact.

What's next for SentryX

Add video deepfake detection with higher accuracy.

Integrate browser extensions or social media plugins for real-time verification while browsing.

Expand the system to multi-language fact-checking.

Build a public dashboard where users can submit content and track trending misinformation.

Built With

• ai/deepware)
• and
• apis
• built-with-react.js
• deepfake
• detection
• flask/fastapi
• for
• google-fact-check-api
• hive
• maestro-ai-agents
• multi-modal
• tailwind-css
• twitter/x-api