Imagine having a platform where you can safely expose corruption in your country without fear of retaliation, and where journalists can analyze thousands of documents to uncover patterns of fraud and abuse. That’s exactly what HaqNow provides – a privacy-first digital sanctuary for truth-seekers across the globe. Named after the Arabic word for “truth” or “right,” this innovative platform allows citizens and journalists to anonymously upload evidence of corruption in over 60 languages, with automatic English translation making these documents accessible to international audiences. Whether it’s government fraud, corporate malfeasance, or public sector corruption, HaqNow creates a secure space for whistleblowers to shine light on wrongdoing while maintaining their anonymity.

What truly sets HaqNow apart is its revolutionary AI-powered Q&A system built on Retrieval-Augmented Generation (RAG) technology. Instead of manually sifting through thousands of complex documents, users can now ask natural language questions like “What corruption cases involve Brazil?” or “What types of government fraud are mentioned?” and get detailed, synthesized answers complete with source attribution and confidence scoring. The system works with documents in all supported languages and uses open-source models processed locally to maintain privacy, ensuring that the fight against corruption doesn’t compromise user security. This intelligent discovery tool transforms how investigators and journalists can analyze massive document dumps, finding relevant information across thousands of files in seconds.

Behind the scenes, HaqNow runs on a sophisticated yet cost-effective architecture that prioritizes both functionality and ethical considerations. The platform uses a dual database system with MySQL for document metadata and PostgreSQL with pgvector for AI embeddings, while being powered by Thaura.AI’s privacy-first LLM and sentence-transformers for local processing. The document pipeline carefully handles uploads through admin approval, OCR text extraction, and chunking before generating embeddings for vector search. With built-in analytics through self-hosted Umami that respects user privacy (no cookies, no stored IPs), HaqNow proves that powerful anti-corruption tools can be both effective and ethically sound. For anyone looking to fight corruption safely and effectively, HaqNow represents a new frontier in digital activism and transparency.

What if you could track how green your city really is? The Greenspace Detection App brings satellite imagery analysis right to your desktop, turning complex data into actionable insights about urban vegetation. This powerful tool automatically downloads and processes satellite imagery from cities worldwide, removing clouds and using advanced NDVI calculations to identify and visualize vegetation with remarkable accuracy. Whether you’re an urban planner, environmental researcher, or just curious about your city’s green cover, this app makes sophisticated environmental analysis accessible to everyone.

The magic happens through a scientifically rigorous approach that processes Sentinel-2 satellite imagery at 10-meter resolution, filtering out clouds and shadows to focus on what matters most – vegetation. Using the NDVI formula (NIR – Red) / (NIR + Red), the app categorizes vegetation into density levels from subtle growth to dense forests. What makes this particularly powerful is the temporal analysis feature, which compares annual vegetation patterns by selecting the best low-cloud scenes each month, then averaging the results to give you comprehensive insights into how your city’s green spaces are changing over time.

Beyond just pretty maps, this desktop application delivers real analytics that can drive meaningful environmental decision-making. With 50+ cities already mapped and predefined boundaries, you can instantly explore vegetation coverage statistics, track changes between years, and identify areas gaining or losing green space. The app handles all the complex satellite processing in the background while providing live progress updates, so you get scientific-grade environmental intelligence without needing to be a remote sensing expert. It’s democratizing environmental data analysis, one city at a time.

https://community.haqnow.com/

https://github.com/main-salman/haqnow.community

If you’ve ever wanted to build a secure, collaborative platform for handling large sets of community documents, community.haqnow is a solid starting point. It’s designed to support bulk uploads, OCR + translation, powerful search, AI-driven Q&A, and team collaboration—all the things you’d want if you’re managing big collections of documents across languages and formats. Think of it as the infrastructure backbone that makes knowledge-sharing easier and more transparent.

Under the hood, this repo comes with all the tooling you need to spin things up quickly. It uses Terraform IaC for Exoscale, helper scripts for deployment, and a Makefile to keep repetitive tasks clean and simple. Deployment is modeled after the main HaqNow platform, so you can bootstrap a server, drop in your app code, and get a working backend/frontend without a headache. Secrets are kept safe in .env files (never committed), while database access is tightly restricted down to your VM’s IP—so you’re not just fast, you’re secure by default.

The architecture itself is straightforward but flexible: a Compute VM running Ubuntu and nginx serves your static site and API endpoints, while Exoscale’s DBaaS handles both MySQL and Postgres needs. Object storage via S3 powers your document buckets, making files easily accessible while keeping the rest of your setup clean and modular. In short, community.haqnow gives you a strong foundation for any project that needs reliable infrastructure for community-driven knowledge sharing.

Big news — we’re getting ready to launch HaqNow.com, a global platform built to shine a light on corruption, wherever it hides. The name “Haq” comes from Arabic, meaning “truth” or “right,” and that’s exactly what we’re about. Anyone — whether you’re a journalist, activist, or just a concerned citizen — can anonymously upload documents in over 60 languages, with instant English translation so the whole world can read and understand. Privacy comes first, and we’ve designed the platform so people can share critical information without risking their safety.

But here’s where it gets even more exciting — we’re introducing a brand-new AI-powered Q&A system that can actually answer questions about the corruption documents on the platform. Wondering “What corruption cases involve Brazil?” or “What types of government fraud are mentioned?” Just ask in plain language and get detailed answers, complete with clickable sources and a confidence score so you know how reliable the response is. And the best part? It’s 100% powered by open-source tech and runs locally for maximum privacy — no big corporate AI peeking at your data.

We’ve packed a lot of innovation into the system: a dual database setup for speed and accuracy, cutting-edge language models from Ollama and sentence-transformers, and smart search capabilities that work across thousands of documents in seconds. Everything from upload to answer generation is designed to make finding the truth faster, easier, and safer. HaqNow is more than a website — it’s a tool for global transparency. We can’t wait to open the doors and see how people around the world will use it to hold power to account.

Ever wondered how to uncover hidden patches of greenery in your city without wrestling with API keys or complex satellite data pipelines? That’s exactly what the greenspace-detection-platform is all about. Born from a clever mix of Next.js, Python, STAC, and Sentinel-2 satellite data (via the Microsoft Planetary Computer), this tool delivers beautifully aligned, key-free vegetation analysis, boosting detection by a whopping 46% compared to standard methods GitHub.

https://github.com/main-salman/greenspace-detection-platform

What’s particularly fun is how it brings everything to life: you pick your city (Toronto’s even on the list!), tweak sensitivity and date range, and watch the platform render a purple-gradient map showing dense forests, parkland, grass, and subtle green hints—all geo-aligned to sub-pixel accuracy and ready to toggle between grayscale, OSM, or satellite views GitHub. Under the hood, sub-pixel precision and smart CRS transformations ensure images align perfectly, while Leaflet, OpenCV, and Rasterio stitch it all together in the frontend and backend.

Whether you’re a hackathon pro or just curious about urban ecology, this is a seriously cool project to play with—or build on. Installation is a breeze with npm install, npm run setup, and then launch with npm run dev GitHub. You’ll be inspecting greenness in your favorite city in just a few clicks—perfect for anyone with an eye for code and a soft spot for nature.