-
-
Research Helper, Happy to Help!
AI Research Helper - Project Description
Inspiration
Research can be overwhelming. As students and professionals, we found ourselves drowning in dozens of browser tabs, scattered Google Docs, and endless articles while trying to piece together coherent research projects. Existing AI tools either required uploading sensitive documents to the cloud or lacked the contextual understanding needed for academic work. We were inspired by Chrome's groundbreaking on-device AI capabilities and saw an opportunity to create a privacy-first research assistant that could analyze documents locally while providing intelligent academic guidance. The goal was to transform chaotic research sessions into organized, structured insights without compromising user privacy.
What it does
AI Research Helper is a Chrome extension that acts as your intelligent research companion. It features a floating sidebar widget that appears on any webpage, providing instant AI-powered content analysis using Chrome's built-in Gemini Nano model, Prompt API, Writer API and Summarizer API. The extension extracts meaningful keywords from academic content, generates research directions and themes, recommends relevant academic databases, and creates structured research outlines. For enhanced functionality, it integrates with Google Docs through OAuth 2.0, allowing deeper document analysis while maintaining read-only access. Everything processes locally on your device, ensuring complete privacy. The extension also includes note-taking capabilities, research progress tracking, and fallback analysis when Chrome AI isn't available.
How we built it
We built the extension using Chrome Manifest V3 with a focus on privacy-first architecture. The core uses Chrome's experimental Gemini Nano AI API for local content analysis, with sophisticated content extraction algorithms that filter out UI elements and focus on meaningful text. We implemented multi-layer content processing: starting with Google Docs API integration via OAuth 2.0, falling back to DOM parsing, and finally using text extraction as a last resort. The extension features a responsive floating sidebar built with vanilla JavaScript and CSS, ensuring minimal performance impact. We created intelligent keyword extraction using context windows and semantic analysis, avoiding common stop words and UI elements. The OAuth implementation follows Chrome extension security best practices, using the identity API with proper token management and user consent flows.
Challenges we ran into
Content Extraction Complexity: Different websites and Google Docs have vastly different DOM structures. We had to build robust content extraction that could handle everything from academic papers to news articles while filtering out navigation, comments, and ads.
OAuth Configuration: Managing consistent extension IDs across development environments proved challenging. The "bad client id" error required implementing proper key management and clear troubleshooting documentation.
Privacy vs Functionality Balance: Providing powerful analysis while keeping everything local required innovative approaches to content processing and careful API selection.
Cross-Platform Compatibility: Ensuring the extension works across different operating systems and Chrome versions while maintaining consistent behavior.
Accomplishments that we're proud of
Privacy-First Innovation: Successfully implemented powerful AI analysis that runs entirely on-device, setting a new standard for privacy-conscious research tools.
Seamless Integration: Created a non-intrusive floating widget that enhances rather than disrupts the browsing experience, working on any website.
Robust Content Analysis: Developed sophisticated algorithms that can intelligently extract meaningful content from complex web pages and documents while filtering noise.
Professional Documentation: Built comprehensive setup guides, privacy policies, and developer documentation that make the extension accessible to both users and contributors.
Open Source Contribution: Created a fully open-source project with proper licensing, security practices, and community guidelines.
Google Docs Integration: Successfully implemented secure OAuth 2.0 integration that enhances functionality while maintaining user control and privacy.
What we learned
Emerging AI Technologies: Gained deep expertise in Chrome's experimental AI APIs and learned how to build applications that gracefully handle cutting-edge, evolving technologies.
Privacy by Design: Understood that privacy isn't just about compliance—it's a competitive advantage and user trust builder that can differentiate products in the market.
Chrome Extension Architecture: Gained experience with Manifest V3 development, security best practices, and the intricacies of building robust browser extensions.
OAuth Implementation: Learned the complexities of secure authentication in browser extensions and the importance of clear developer documentation.
User Experience Design: Discovered how to create helpful tools that enhance workflows without being intrusive or overwhelming.
Open Source Community Building: Understood the importance of comprehensive documentation, clear contribution guidelines, and transparent development practices.
What's next for Research Helper AI (RHAI)
Enhanced AI Capabilities: Integrate more advanced local AI models as they become available, including specialized academic and research-focused language models.
Collaborative Features: Add team research capabilities with shared workspaces, collaborative annotations, and research project management tools.
Academic Database Integration: Direct integration with scholarly databases like PubMed, IEEE Xplore, and JSTOR for seamless source verification and citation management.
Citation Management: Built-in citation formatting, bibliography generation, and integration with tools like Zotero and Mendeley.
Mobile Companion: Develop companion mobile apps for research on-the-go while maintaining the privacy-first approach.
Advanced Analytics: Research progress tracking, productivity insights, and personalized research recommendations based on user patterns.
Chrome Web Store Publication: Complete the verification process and publish to make the extension accessible to the broader academic community.
Community Contributions: Build a community of academic users and developers who can contribute features, language packs, and domain-specific research tools.
Elevator Pitch
"Turn research chaos into organized insights using Chrome's privacy-first on-device AI; Analyze documents, generate academic directions, and structure your research without sending data to the cloud."
Key Features
- Local AI Processing - Uses Chrome's built-in Gemini Nano model
- Privacy-First Design - No data leaves your device
- Google Docs Integration - Secure OAuth 2.0 document analysis
- Smart Content Analysis - Intelligent keyword extraction and research suggestions
- Research Organization - Note-taking, progress tracking, and structured outlines
- Universal Compatibility - Works on any website with floating sidebar widget
- Academic Focus - Tailored for students, researchers, and professionals
- Open Source - Transparent, community-driven development
Our vision is to make AI Research Helper the go-to research companion for students, academics, and professionals worldwide, proving that powerful AI tools can respect user privacy while enhancing intellectual work.
Log in or sign up for Devpost to join the conversation.