RateMyGaucho (Chrome Extension)

RateMyGaucho enhances UCSB GOLD to 1,200+ students per month by showing professor ratings and course data directly on course result pages. It works offline and preserves your privacy by using a local dataset.

Demo Video: https://www.youtube.com/watch?v=fl7-icSAves

Novel Features

• Offline & Private: Works without an internet connection and never sends your data to external servers. All data lives in a single packaged CSV (courses_final_enrollment.csv).
• Gaucho-Themed Ratings: Displays professor ratings with custom, partially-filled Gaucho star icons for at-a-glance assessments.
• In-Depth Course Data: Goes beyond professor ratings to show course-specific details like grading trends, enrollment history, and recent student reviews.
• Visual Summaries: Inline bar charts translate grade distributions and historic enrollment snapshots into easy-to-interpret visuals.
• Smart Review Filtering: Intelligently filters reviews to show only those relevant to the specific instructor, providing more accurate insights.
• UCSB Plat Integration: Provides direct links to professor profiles and curriculum pages on UCSB Plat.

Installation

1. Download the latest release from the Releases page.
2. Extract the ZIP file.
3. Open Chrome, go to chrome://extensions, and enable "Developer mode".
4. Click "Load unpacked" and select the extracted folder.

Building

To create a distributable ZIP file, run the appropriate script for your OS:

• Windows: ./scripts/package.ps1
• macOS/Linux: bash ./scripts/package.sh

The build script bundles the unified dataset and content script into dist/RateMyGaucho.zip, which is ready to upload to the Chrome Web Store.

Data Source

The extension ships with courses_final_enrollment.csv as its only data source. Every rating, review, and course record comes from this file, ensuring consistent results across the UI. When updating data, replace this CSV and rebuild the package.

---

Backend Architecture

The backend/ directory contains a production-grade Python API that powers search, analytics, and data ingestion.

Tech Stack

Layer	Technology
API	Python 3.12 · FastAPI · Pydantic v2
Database	PostgreSQL 16 (SQLAlchemy ORM, connection pooling)
Search	Elasticsearch 8.17 (full-text, fuzzy matching, filters)
Cache	Redis 7 (cache-aside, 300 s TTL, 128 MB LRU)
Messaging	Apache Kafka 3.8 KRaft (no ZooKeeper)
Observability	OpenTelemetry → Jaeger (distributed tracing)
Orchestration	Docker Compose (all services with health checks)

Key Components

• ETL Pipeline (app/etl.py) — Reads the CSV, normalises into Department → Professor → Course → Review tables in Postgres, bulk-indexes into Elasticsearch, and publishes Kafka events.
• Elasticsearch Service (app/es_service.py) — Index management, multi-match search with fuzziness, department/GPA/professor filters.
• Redis Cache (app/cache.py) — JSON cache-aside pattern reducing median latency from ~650 ms to ~180 ms.
• Kafka Producer / Consumer (app/kafka_producer.py, app/kafka_consumer.py) — Event-driven pipeline; the consumer upserts into Postgres + ES on course.created / course.updated events.
• REST API (app/routes.py) — /search/courses, /search/reviews, /courses, /analytics, /health
• Tracing (app/tracing.py) — Auto-instruments FastAPI, SQLAlchemy, Redis, and Elasticsearch via OTLP → Jaeger.

Quick Start

# From the repo root
docker compose up --build -d

# API at http://localhost:8000
# Jaeger UI at http://localhost:16686
# Elasticsearch at http://localhost:9200

API Examples

# Health check
curl http://localhost:8000/api/v1/health

# Search courses
curl -X POST http://localhost:8000/api/v1/search/courses \
  -H 'Content-Type: application/json' \
  -d '{"query": "machine learning", "department": "CMPSC"}'

# Analytics
curl http://localhost:8000/api/v1/analytics

Running Tests

cd backend
pip install -r requirements.txt
pytest

Contributing

Contributions are welcome! Please fork the repository and submit a pull request.

License

MIT © 2025