Noor Ul Hassan@noorulhassan304

ConnectSense: Bridging South Asia's Digital Divide ConnectSense addresses a critical challenge across South Asia, where over 900 million people in rural and remote communities lack reliable internet due to challenging geography, severe weather, limited budgets, and complex regulations. This digital exclusion impacts education, healthcare, and economic opportunities in the region's most vulnerable communities. Designed specifically for non-technical stakeholders like government officials, school administrators, and healthcare providers, ConnectSense is an AI-powered connectivity advisor that transforms complex telecommunications decisions into accessible guidance. The system evaluates geographical conditions, assesses appropriate technologies from fiber to satellite, optimizes budgets, and navigates country-specific regulations to deliver customized connectivity solutions in plain language. Built on a Python-based architecture using FastAPI, LlamaIndex, and FAISS vector database technology, ConnectSense processes region-specific connectivity knowledge through multiple AI models including Groq, and Gemini. Its Streamlit-powered interface offers an intuitive chat experience that maintains conversation history for iterative planning. ConnectSense enables real-world impact across diverse scenarios: helping Nepalese school principals identify satellite options within budget constraints, supporting Bangladeshi health officials in deploying weather-resistant networks for telemedicine, guiding Pakistani village councils through licensing requirements, and assisting Indian administrators with phased connectivity planning. By democratizing access to telecommunications expertise, ConnectSense empowers communities to build sustainable digital infrastructure and create pathways to opportunity in South Asia's most underserved regions.

Imagine being Amna from a rural area in pakistan, facing a legal challenge but surrounded by barriers. The nearest legal aid center is 45 kilometers away, legal fees exceed her monthly income, and she fears social stigma. Her story echoes across Pakistan, where 85% of women face similar barriers in accessing legal support. The numbers tell a heartbreaking story - only 1 in 3 women experiencing abuse seeks legal help, 68% of rural cases go unreported, and with legal consultation costing PKR 10,000-15,000, justice remains out of reach for many. When you add that only 39% of women have internet access and the female literacy rate is 42.7%, the challenge becomes even more daunting. Project Hifazat steps in as a digital guardian, combining IBM's Granite AI technology with deep cultural understanding. At its heart lies the IBM Granite-3-8b-instruct model, carefully molded to understand not just legal language, but Pakistani society's nuances. Paired with the Granite-embedding:278m multilingual model, our system breaks down all barriers, offering support to pakistani females. Unlike typical legal AI systems, Project Hifazat acts as a culturally-aware legal companion. When Amna shared her property rights case, our system immediately provided legal information accordingly, offered confidential guidance, connected her with support services (locally), and gave step-by-step documentation guidance. The magic happens through our innovative use of IBM's Granite models. The 8B language model brings legal expertise with cultural sensitivity, while the multilingual embedding model ensures women can express concerns in their preferred language. This isn't just translation - it's about understanding context, emotion, and urgency. With this, we'll help rural women understand property rights, working women address harassment, and mothers navigate family law. Looking ahead, we'll work on court system integration, expanding language support, and enhancing mobile accessibility.