Inspiration

The idea for Cognito was born from a shared, modern frustration: the fleeting nature of truth online. We've all been in debates where an article's headline is silently edited or a controversial post is quietly deleted. We were inspired by a simple but powerful question: What if we could create a permanent, unchangeable record of what we read? Our goal wasn't just to analyze content, but to anchor that analysis to an immutable source of truth, giving every citizen a tool to fight misinformation.

What it does

Cognito is an AI-powered "Critical Thinking Co-Pilot" that helps users analyze and verify online content. Here's what it does:

  • Analyzes Any Article: A user submits any URL, and our AI backend provides a comprehensive report on the article's sentiment, potential biases, and key factual claims.
  • Creates a Verifiable Timestamp: With a single click, Cognito creates a permanent, immutable record of the analysis on the Algorand blockchain. This acts as a digital notary, proving the article's content existed in that specific form at that exact time.
  • Saves and Manages Reports: Users can create an account to save their analysis history to a personal, searchable dashboard with full filter and delete functionality.
  • Makes Information Accessible: The app can generate an audio version of any report using ElevenLabs' AI and allows users to export a professional PDF for offline use or sharing.

How we built it

Cognito is a full-stack web application built with a focus on modern, powerful, and scalable technologies. The entire development process was orchestrated within bolt.new.

  • Frontend: React, TypeScript, Tailwind CSS
  • Backend: Node.js, Express
  • Database & Auth: Supabase (PostgreSQL, Auth, Triggers)
  • AI Analysis: Google Gemini API
  • Blockchain: Algorand (via Algorand JavaScript SDK)
  • Audio Generation: ElevenLabs API
  • PDF Generation: jspdf & jspdf-autotable
  • Deployment: Netlify

Challenges we ran into

Every ambitious project has its hurdles, and these were our crucial learning moments:

  1. Navigating Evolving APIs: Our AI development tool, bolt.new, occasionally generated code based on older documentation for cutting-edge services like Algorand and Google Gemini. We overcame this by manually consulting the latest official API documentation, identifying the deprecated methods, and refactoring the code to be fully compliant.
  2. State Management in the Dashboard: Getting the "Save Report" feature to update the UI instantly without a page refresh was surprisingly tricky. We had to manually trace the data flow in React and refactor our state and effect hooks to resolve a subtle race condition.
  3. Wrangling the AI's Creativity: The AI would sometimes wrap its perfect JSON response in helpful, human-readable sentences, which broke our backend parser. We solved this by building a more resilient parser that surgically extracts only the JSON block from the AI's raw text response.

Accomplishments that we're proud of

  • A True Multi-API Symphony: We successfully integrated five distinct, complex APIs (Supabase, Algorand, Google AI, ElevenLabs, PDF-generation) into a single, cohesive user experience.
  • Innovative Use of Blockchain: We are incredibly proud of using Algorand for a powerful, non-financial use case. Creating a "Trust Engine" to fight misinformation demonstrates the true potential of blockchain technology beyond currency.
  • Building a Polished, Full-Stack Product: In a short time, we went beyond a proof-of-concept to build a feature-complete application with secure authentication, a fully functional CRUD dashboard, and a professional, intuitive user interface.
  • Attention to Detail: Features like the audio playback and the cleanly formatted PDF export show a commitment to polish and user experience that we believe sets Cognito apart.

What we learned

  • Blockchain is a Trust Engine: We learned that the most powerful application of blockchain isn't just finance, but its ability to create a single, immutable source of truth that anyone can independently verify.
  • The Developer is the Final Validator: AI development tools are incredible accelerators, but this project taught us that the developer's most important role is to be the "human in the loop"—verifying, testing, and refining AI-generated code to meet production standards.
  • The Power of Good Architecture: Taking the time to properly set up our Supabase database with a separate profiles table and an automated trigger was a key lesson. A solid foundation made implementing complex features much easier down the line.

What's next for Cognito

We believe Cognito is just scratching the surface of what's possible. Our vision for the future includes:

  • Browser Extensions: Develop extensions for Chrome and Firefox to allow users to analyze and timestamp articles with a single click, without ever leaving the page they're on.
  • The "Cognito Archive": Create a public, searchable database of all timestamped analyses, allowing users to track how specific articles or topics evolve over time.
  • Version Comparison: Introduce a feature that allows a user to analyze the same URL twice and see a "diff" that highlights exactly what text was added, changed, or removed between the two analyses.
Share this project:

Updates

Submission history