Decluttr

Decluttr is a Chrome extension that serves as your personal productivity assistant during the job search.

Comment

Decluttr - AI-Powered Productivity Assistant

Inspiration

As computer science students preparing for Summer 2026 internships, we found ourselves drowning in email chaos. Our inboxes were flooded with application confirmations from 100+ companies, two-factor codes mixed with critical recruiter messages, promotional spam from job boards, and time-sensitive updates buried under clutter. Simultaneously, crafting personalized LinkedIn messages for every recruiter felt overwhelming while juggling coursework and applications. We built Decluttr to solve our own problem—and realized every student hunting for internships faces the same struggle.

What it does

Decluttr is a Chrome extension that serves as your personal productivity assistant during the job search. It features three core modules:

AI Email Assistant:

  • Smart Categorization: Automatically separates emails into Primary and Promotions tabs
  • AI-Powered Summaries: Generates 2-3 sentence summaries of lengthy emails using GPT-4o-mini
  • Intelligent Unsubscribe Detection: Identifies unsubscribe links and provides one-click bulk unsubscribe
  • Bulk Email Management: Select multiple senders, unsubscribe from all, and optionally trash all their emails

LinkedIn Outreach Automation:

  • Recruiter Discovery: Search LinkedIn for recruiters using natural language queries
  • Profile Parsing: Automatically extracts name, title, company, and location from search results
  • AI Message Generation: Creates personalized, professional connection messages tailored to each recruiter
  • One-Click Workflow: Search → Select Profile → Generate Message → Copy → Send (under 30 seconds)

Focus Mode:

  • Block distracting websites during study sessions
  • Pomodoro timer integration
  • Productivity analytics and time tracking

How we built it

We built Decluttr as a Chrome extension with a Node.js backend:

Frontend:

  • Vanilla JavaScript (ES6 modules) for clean code organization
  • Chrome Extension APIs (chrome.identity for OAuth, chrome.storage.local for caching)
  • HTML5 + CSS3 with glassmorphism design (backdrop-filter blur effects)
  • Custom SVG logo with pure CSS animations

Backend:

  • Node.js with Express.js REST API
  • OpenAI API (GPT-4o-mini) for email summaries and LinkedIn messages
  • SerpAPI for LinkedIn profile discovery via Google search
  • Gmail API integration with OAuth 2.0 for email access

Architecture:

  • Repository pattern for centralized email data management
  • Service layer for business logic (Gmail, OpenAI, parsing, unsubscribe)
  • Controller pattern for UI and event orchestration
  • Smart caching system to avoid redundant AI calls

Challenges we ran into

Setting up Google Cloud Platform

Configuring Gmail API OAuth for Chrome Extensions was confusing at first. Chrome Extensions require special redirect URIs that must match exactly in Google Cloud Console, and the documentation wasn't very clear. We spent a lot of time figuring out the correct OAuth flow and troubleshooting redirect_uri_mismatch errors before we finally got authentication working.

Choosing the right APIs

We had to research and compare multiple APIs to find the best fit for our project. For LinkedIn search, we initially chose Google's Custom Search API but it got deprecated mid-development. We had to quickly pivot to SerpAPI. For AI features, we evaluated different OpenAI models to balance cost and quality, eventually choosing GPT-4o-mini for its speed and affordability.

Balancing school assignments

Building Decluttr while juggling midterms, assignments, and other coursework was tough. We had to carefully manage our time, work in short sprints between classes, and prioritize features to stay on track. There were definitely late nights debugging code right after finishing problem sets.

Accomplishments

Fully functional Chrome extension with seamless integrations

We successfully integrated Gmail API, OpenAI API, and SerpAPI into a polished, production-ready extension. The three-click LinkedIn workflow (Search → Select → Copy) takes under 30 seconds per connection.

Clean architecture without frameworks

We built maintainable, modular code using vanilla JavaScript with proper separation of concerns (Repository, Service, Controller patterns). This proved frameworks aren't required for good architecture and allowed us to pivot APIs in 20 minutes when needed.

Measurable real-world impact

Students save ~95 minutes per day on email management and LinkedIn outreach. Over a 100-day internship application season, that's 158 hours saved (nearly 4 full work weeks) that can be spent coding, learning, or applying to more companies.

What we learned

Throughout this project, we learned how to build production-quality browser extensions from scratch. We gained hands-on experience with OAuth 2.0 authentication flows, MIME email parsing standards, and prompt engineering for AI generation. We discovered that good software architecture is more important than using the latest framework. When Google deprecated our search API mid-development, our abstraction layer let us pivot in 20 minutes instead of days. We learned to handle CORS restrictions creatively, implement graceful fallbacks when APIs fail, and write error messages that actually help users fix problems. Most importantly, we learned that building tools to solve our own pain points results in better products, we use Decluttr every single day for our own internship hunt.

What's next for Decluttr

Immediate improvements:

  • Enhanced error recovery with exponential backoff
  • Rate limiting and request queuing
  • Unit tests for email parsing logic
  • Analytics dashboard for usage statistics

Feature expansion:

  • Email templates for common recruiter responses
  • Smart filters that learn from user behavior
  • Enhanced LinkedIn search with filters (company size, location, industry)
  • Connection request tracking (sent, accepted, pending)
  • Full Focus Mode implementation with website blocking and Pomodoro timers

Growth strategy:

  • Beta launch with Carleton University students (Spring 2026)
  • Expand to other Canadian universities (Fall 2026)
  • Open beta to students globally (Winter 2027)
  • Premium tier with advanced AI features and unlimited searches
  • University partnerships for institution-wide student access

Vision: Eventually, we want Decluttr to become the go-to productivity platform for students navigating job searches, with integrations for Handshake, LinkedIn Jobs, Indeed, and applicant tracking across all platforms. We're building the career assistant we wish we had.

Built With

Share this project:

Updates