Problem Statement: People with motor disabilities struggle to use digital devices. Voice commands are slow, and existing hardware solutions are expensive. 💡 Solution: An AI-powered eye-tracking cursor using IBM Granite Vision Transformer to enable hands-free control of PCs and smartphones with just eye movements and blinks. 🚀 Key Features: ✅ Gaze-Based Cursor Movement – Move the cursor by looking around. ✅ Blink to Click – Short blink for left-click, long blink for right-click. ✅ Gaze Scrolling – Look up/down to scroll pages and reels. ✅ Real-Time AI Adaptation – Model learns user-specific gaze patterns. ✅ No Extra Hardware Needed – Works with a standard webcam.

PoliBot is an advanced global crisis negotiation simulator built using Python and Streamlit. The core idea is to model complex international relations and decision-making during various global emergencies (like climate collapse, pandemics, or energy crises) using autonomous AI agents. Each agent represents a specific country (e.g., USA, China, India, EU, Pakistan), equipped with a unique profile detailing its strengths, weaknesses, and national interests. Leveraging the Groq API and the Llama3 language model, these agents analyze the ongoing scenario, consult their recent memory of interactions, and decide on diplomatic actions—such as proposing deals, building alliances, or raising concerns—to best serve their country's objectives. The simulation dynamically tracks key global metrics (Peace Index, Economic Growth, CO₂ Emissions, etc.) and visualizes the evolving diplomatic network and relationships between nations using NetworkX and Matplotlib. Users can configure the scenario, participating nations, simulation speed, and other parameters, observing how AI-driven diplomacy unfolds turn by turn. The project demonstrates the potential of LLMs in simulating complex multi-agent systems and geopolitical dynamics.