July 11th, 2025

Automating User Feedback Analysis for Al-Quran App

This blog post introduces Greentech Apps Foundation's automated Comment Analysis Project for the Al-Quran app. By using Python scripts, machine learning, and Slack integration, the system classifies daily app reviews into categories like bug reports, feature requests, and content issues. It automatically stores relevant data in a database and sends weekly summaries to the team. This automation helps GTAF prioritize product updates and respond quickly to user needs, streamlining feedback analysis across platforms in both English and Arabic. The project is part of GTAF’s broader R&D efforts to use AI in building user-driven Islamic apps.

Published: 05 March 2024

Author: Mohammad Galib Shams, Nabil Mosharraf Hossain

At Greentech Apps Foundation, we constantly strive to connect with our users, understand their needs, and improve our products based on their feedback. One of our key projects toward this goal is the Comment Analysis Pipeline—a backend automation tool that helps us monitor, classify, and act on user reviews from the app stores.

The Challenge

With tens of thousands of user comments pouring in across platforms, manually reviewing each one isn’t scalable. Important feedback—like bug reports, feature requests, or content errors—can get buried among general praise or unrelated messages. We needed a way to automatically detect, classify, and summarize what users are saying about our Al-Quran app, every single day.

The Solution: Comment Analysis Pipeline

We built a lightweight Python-based pipeline that runs daily and performs the following tasks:

1. Load Fresh Reviews

Using the Data() class, the script pulls in all reviews posted in the last 24 hours.

review_last_day = data.get_data_last_day()

2. Classify Key Issues

It uses topic classification to tag reviews that mention problems, such as bugs, content errors, or feature/content requests.

problems_last_day = data.get_topic_classification(problems_last_day=review_last_day[review_last_day["type"]=="problem"])

3. Save to Database

Classified reviews are saved to our PostgreSQL database for long-term analysis, tagged with additional metadata like topics and explanation.

data.save_to_db(problems_last_day[...], table_name="AL_QURAN", if_exists="append")

4. Weekly Summary Report

If it’s the end of the week, the system sends a Slack summary to the internal team—highlighting the most common issues users faced, based on classified topics.

if result == True:
    data.send_summarization()

Why It Matters

This pipeline helps us prioritize product improvements based on real user needs. It also makes our internal team more responsive, ensuring that valuable feedback doesn’t go unnoticed. From catching broken ayah audio to understanding high-demand features like tafsir or dark mode, this system keeps us closely aligned with our global user base.

What’s Next

We’re exploring:

Integrating multilingual support to handle reviews in non-English languages more accurately
Linking feedback trends to product development sprints
Building dashboards for live comment insights

Final Thoughts

The Comment Analysis Project is a small but powerful step in using AI for user-centric Islamic tech. By turning raw feedback into actionable insights, we’re not just building apps—we’re building trust.

🔍 Want to contribute or learn more about our R&D efforts? Visit gtaf.org or follow us on LinkedIn.

Share the Knowledge