One of the biggest challenges in my work is making sense of scattered user feedback. People talk about mental health apps everywhere, but finding structured insights is tough. That’s why I built RedCrawler, a Python-based tool that scrapes and analyzes Reddit discussions about mental health apps. With help from AI during development, and by analyzing the JSON outputs in Claude Code, I turned a raw idea into a practical research engine.

What RedCrawler Does

At its core, RedCrawler is designed to answer three big questions:

1. How do people use mental health apps? Usage frequency, contexts, and patterns.
2. Which apps do they recommend? Popularity, sentiment, and recommendation networks.
3. For which conditions do they use apps? Condition–app mappings and reported effectiveness

It works by scraping posts and comments across 20+ mental health subreddits, running advanced text analysis such as sentiment scoring, keyword extraction, and pattern detection, and generating visualizations like network graphs, sentiment charts, and word clouds.

The result: structured, real-world insights into how people perceive and use apps like Calm, Headspace, BetterHelp, and Daylio.

Building It with AI

I developed RedCrawler in Python, but AI was my co-pilot throughout (Cursor to be exact). It helped me design the architecture, debug code, and even draft tests and documentation. Instead of getting bogged down in technical hurdles, I could focus on the research outcomes I wanted to unlock.

The real breakthrough came when I combined RedCrawler with Claude Code. By loading the JSON output into Claude, I could query thousands of Reddit posts in natural language, surfacing trends, anomalies, and correlations in hours instead of days. This workflow made data analysis feel less like heavy lifting and more like having an interactive research partner.

Why This Matters for My Work

For me, RedCrawler is not just a technical project. It is a business and research tool. It saves hours of manual effort, ensures more reliable reporting, and scales easily as new apps, subreddits, or conditions emerge.

Most importantly, it shows how AI can amplify expertise. By combining automation (Python), unstructured data (Reddit), and AI-powered analysis (Claude Code), I created a system that transforms messy online conversations into actionable insights.

This has direct relevance to my work in mental health and technology: it helps me better understand user needs, evaluate digital health trends, and support decision-making with real evidence.

Final Thought

RedCrawler proves that the future of work is not about replacing humans with AI. It is about building the right tools to multiply our impact. With the right mix of automation and intelligence, we can turn the overwhelming noise of online discussions into clear insights that drive innovation and, ultimately, help more people.