Smart Research Assistant

Transform research forever with AI summaries, smart questions, and content highlighting—all running locally on Chrome's built-in AI for complete privacy and zero cost.

Comment

Inspiration

  • Research workflows today are broken and fragmented.
  • People read articles passively without extracting key insights.
  • They miss important follow-up questions that could deepen their understanding.
  • Research progress is often lost across sessions.
  • Many rely on external AI services that compromise privacy while costing money.
  • Chrome's built-in AI APIs could solve these problems by bringing AI directly into the browser where research happens.

What it does

  • Transforms passive web browsing into active, intelligent research.
  • Uses Chrome's Summarizer API to extract key insights from webpages.
  • Generates thoughtful research questions with the Prompt API.
  • Provides content rewriting using the Rewriter API for different perspectives.
  • Features smart content highlighting that shows what AI found important.
  • Includes comprehensive export to Markdown.
  • Offers a complete research note management system.
  • Runs entirely locally on the device with full privacy.

How we built it

  • Built with vanilla JavaScript and Chrome Extension Manifest V3.
  • Core AI processing runs in the content script with Chrome APIs:
    • Summarizer API for summaries.
    • Prompt API for research questions.
    • Rewriter API for content enhancement.
  • Background service worker manages communication between components.
  • Popup UI built with CSS Grid, smooth animations, and intuitive UX.
  • Local storage used for research notes, no external dependencies.
  • Works offline once AI models are downloaded.

Challenges we ran into

  • Chrome’s AI APIs were in early preview with limited documentation.
  • API formats changed during development.
  • Needed robust error handling for different Chrome installations.
  • Designed fallback systems when models weren’t downloaded.
  • Extracting clean content from diverse websites required advanced parsing.
  • Integrating multiple AI APIs while keeping performance high was difficult.

Accomplishments we’re proud of

  • Integrated multiple Chrome AI APIs into a cohesive workflow.
  • Achieved complete privacy with zero data collection.
  • Built intelligent content highlighting linking summaries to source text.
  • Designed a comprehensive export system for publication-ready research.
  • Added keyboard shortcuts for power users.
  • Created a polished UI that rivals commercial applications.
  • Maintained strong performance and security.

What we learned

  • Local AI processing is powerful for privacy and performance.
  • Thoughtful UX design is critical when integrating AI features.
  • Robust error handling is essential with preview APIs.
  • Combining multiple AI APIs creates emergent capabilities more valuable than the sum of their parts.

What’s next for Smart Research Assistant

  • Add multi-language support with Chrome's Translator API.
  • Implement research theme categorization with automatic tagging.
  • Integrate citation generation and source verification.
  • Expand export formats to PDF and DOCX.
  • Add collaborative research features.
  • Integrate with Notion and Obsidian.
  • Provide advanced analytics to reveal research patterns.
  • Build a comprehensive research ecosystem to transform how people discover, process, and organize information online.

Built With

Share this project:

Updates