Scholarship Finder

AI-powered system that monitors scholarships, detects changes, and sends actionable email recommendations automatically.

Comment 1

ScholarSync AI

AI-powered scholarship monitoring system that detects changes and sends actionable email recommendations automatically.

Overview

ScholarSync AI is an end-to-end automation pipeline that monitors scholarship websites, detects meaningful changes, and notifies users with clear, actionable email recommendations.

Instead of manually checking multiple sources, the system continuously tracks updates and converts raw web changes into structured insights delivered via email.

Features

  • Automated browsing of scholarship websites
  • Detection of new, updated, or removed scholarships
  • Change severity classification (LOW / MEDIUM / HIGH)
  • AI-generated email notifications
  • Actionable recommendations per change
  • Modular, agent-based architecture
  • Schema-validated, production-safe AI outputs

Architecture

The system is built as a modular pipeline combining automation and AI reasoning: Browser Use ↓ Lovable (Structuring + Change Detection) ↓ n8n (Workflow Orchestration) ↓ AI Email Agent (MCP Pattern) ↓ Email Notification

MCP Design (Model–Context–Prompt)

The Email Agent follows the MCP pattern:

  • Model
    Generates professional email content and recommendations.

  • Context
    Structured scholarship change data (name, amount, deadline, severity, source).

  • Prompt
    Enforces tone, structure, recommendation logic, and JSON-only output.

  • JSON Schema
    Guarantees deterministic and valid outputs for downstream automation.

How It Works

  1. Browser Use automatically navigates scholarship websites and extracts data.
  2. Lovable structures raw content and compares it against previous snapshots.
  3. n8n filters only changed scholarships and passes structured context to the Email Agent.
  4. The Email Agent drafts a clear summary and recommended actions.
  5. The Email node sends the final notification automatically.

Example Output

Each notification email includes:

  • A summary of detected changes
  • A list of new or updated scholarships
  • Direct source links
  • Recommended next actions (review, verify, monitor)

This ensures recipients understand both what changed and what to do next.

Challenges and Solutions

Unreliable AI output formatting
Solved using strict JSON Schemas and JSON-only prompts.

Alert fatigue
Solved by classifying change severity and filtering low-impact updates.

AI overreach
Solved by limiting the model to recommendations only; execution remains in n8n.

What I Learned

  • How to safely integrate LLMs into automated workflows
  • The importance of schema validation for AI reliability
  • How MCP improves explainability and auditability
  • How orchestration tools turn AI agents into production systems

Built With

  • Browser Use – Automated web navigation and extraction
  • Lovable – AI structuring and reasoning
  • n8n – Workflow automation and orchestration
  • Large Language Models (LLMs) – Email drafting and recommendations
  • JSON Schema – Structured, validated AI outputs
  • SMTP / Gmail – Email delivery

Future Improvements

  • Dashboard for historical change tracking
  • Slack or webhook notifications
  • Multi-language email support
  • User-defined alert rules
  • Database-backed scholarship history
Share this project:

Updates