Context

I've always loved working in cafés. The smell of coffee, the quiet hum of people talking, and the mix of focus and warmth that makes it easier to get things done. But I kept running into the same issue: it's surprisingly hard to find cafés that actually make good workspaces. Some have weak Wi-Fi, others lack outlets, and reviews online rarely mention those details. That frustration, combined with my curiosity as a designer and developer, inspired me to create Lattelink - a project that combines data scraping, sentiment analysis, and design thinking to help people discover reliable, laptop-friendly cafés.

The Problem

Finding a good café to work from shouldn't require endless scrolling through inconsistent reviews and outdated blog posts. Most café listings scatter important details like Wi-Fi quality, outlet availability, and seating comfort across different platforms, making it difficult to compare options or trust the information. Cities change quickly, too, so even the best curated lists become outdated. I wanted to centralize and refresh that data automatically, turning messy, subjective reviews into a clear, comparable "workability" score.

The Solution

I built an end-to-end system that brings together both engineering and design. A Python scraper collects café data from Google Places, then runs VADER and TextBlob sentiment analysis to interpret how reviewers describe Wi-Fi, outlets, noise, and seating. Those values feed into a custom "Holistic Workability Index," which balances practicality and ambience using Bayesian smoothing. A Node/Express backend structures and serves the data through REST APIs, while a Next.js frontend displays an interactive ranked map with café cards and amenity breakdowns. The setup is modular and ready for new sources like Yelp or Foursquare.

The Result

Lattelink now (locally) serves as a living, curated guide to work-friendly cafés, complete with transparent metrics, confidence scores, and tags that capture each spot's character. It's already made my search for study spaces faster and more enjoyable, and beta users have shared the same experience in their own cities. Beyond being useful, the project strengthened my data engineering and system design skills while letting me build something that genuinely improves my day-to-day life - and hopefully, the lives of others who find comfort and focus in a good cup of coffee.