Inspiration
What it does
How we built it
Challenges we ran into
Accomplishments that we're proud of
What we learned
What's next for EchoTrace
About EchoTrace
Inspiration
Misinformation is one of the biggest challenges in today's digital world. We noticed that misinformation doesn't always begin as an entirely false statement. Sometimes, a factual claim can gradually become misleading through exaggeration, missing context, or small changes in wording as it gets repeated and shared.
Most fact-checking tools focus on determining whether a statement is true or false. We wanted to explore a different question: How can information change over time, and how can small distortions eventually create a completely different narrative?
That question inspired us to create EchoTrace, an AI-powered application that helps users visualize how information could evolve across multiple generations of retelling.
What It Does
EchoTrace allows users to enter an original claim and explore how it might change through repeated communication.
Using artificial intelligence, the application generates hypothetical variations of the original statement and organizes them into an interactive branching graph. Each branch represents a possible transformation, such as exaggeration, omission, altered certainty, or a change in context.
Users can:
- Enter a claim and customize the number of generations and branches.
- Explore an interactive graph showing how statements could evolve.
- Follow individual branches to observe how small changes accumulate.
- Select statements to examine changes in wording and meaning.
- Distinguish documented, inferred, and insufficiently supported relationships through visual indicators.
- Investigate claims using external research sources.
Importantly, EchoTrace distinguishes AI-generated hypothetical transformations from verified real-world information spread. Our goal is not to claim that a particular misinformation pathway actually occurred, but to demonstrate how these distortions could happen.
How We Built It
We developed EchoTrace as a full-stack web application using several technologies:
- TypeScript: For application logic and structured data handling.
- React: For building the interactive user interface.
- Next.js: For connecting the frontend with backend API routes.
- CSS: For designing the application's visual layout and user experience.
- Groq API: For generating AI-powered hypothetical transformations and supporting claim analysis.
- Tavily API: For retrieving external research sources related to user-submitted claims.
- React Flow: For visualizing the relationships between statements as an interactive network.
Our prediction system takes a user's original claim and generates possible mutations across multiple generations. Each new statement develops from its immediate parent rather than directly from the original claim.
We then process the AI-generated results into connected nodes and edges, allowing users to explore how meaning changes throughout the graph.
We also incorporated research functionality to help users examine supporting information rather than relying exclusively on AI-generated content.
Challenges We Faced
One of our biggest challenges was designing a system that could generate meaningful transformations while preserving a logical relationship between statements.
We wanted each generation to build upon the previous one, creating a realistic illustration of how information might gradually become distorted. This required carefully structuring our AI instructions and organizing the resulting statements into a branching hierarchy.
Another challenge was integrating multiple technologies into a single application. Connecting the Groq API, research functionality, backend processing, and interactive graph visualization required debugging and testing.
We also encountered deployment difficulties involving environment variables, API requests, and GitHub's secret protection. These experiences reinforced the importance of securing API keys, managing version control, and understanding the differences between local development and production environments.
Finally, we had to consider how to present AI-generated information responsibly. We worked to make clear that predicted transformations are hypothetical rather than documented events.
What We Learned
Building EchoTrace helped us strengthen our understanding of full-stack web development, API integration, and AI-powered applications.
We learned how to structure a React and Next.js project, communicate with external APIs, process structured AI responses, and transform complex data into an interactive visualization.
We also gained practical experience with Git, GitHub, deployment workflows, debugging, and handling sensitive environment variables.
Most importantly, we learned that AI-generated predictions must be presented carefully and transparently. An AI model can illustrate how misinformation might develop, but those predictions should not be confused with verified historical evidence.
What's Next for EchoTrace
In the future, we hope to expand EchoTrace by improving the accuracy and consistency of its AI-generated transformations, strengthening its research and evidence evaluation features, and making the graph visualization even more intuitive.
We would also like to explore ways to analyze documented examples of misinformation spreading across different sources, while maintaining a clear distinction between verified relationships and hypothetical predictions.
Our long-term vision is to make EchoTrace an educational tool that helps students, researchers, and everyday internet users better understand how information changes and why critical thinking matters.
EchoTrace is more than a misinformation detector. It is a way to visualize how information can evolve—and why understanding that evolution is essential.
Log in or sign up for Devpost to join the conversation.