Hi, I'mFahim Shakil

Published research paper on arXiv presenting a machine learning-based approach to automatically detect auto-rickshaws in traffic images for smart city applications in Bangladesh. Developed a real-time object detection system using YOLOv8 model, achieving mAP50 of 83.44% and precision/recall above 78%. The system performs reliably in complex traffic and low-light conditions, enabling data-driven traffic management decisions. Prepared and publicly released an annotated dataset of 1,730 images captured under various traffic conditions.

Comprehensive architectural visualization platform developed for Google DeepMind sponsored nano banana Kaggle contest. Transforms 2D floor plans into photorealistic, interactive 3D interior visualizations using Google's Gemini 2.5 Flash AI with 95% room detection accuracy. Features intelligent room detection, cost estimation with country-specific pricing, and interactive 3D navigation.

Developed for KUET BitFest Hackathon 2025, an AI-powered platform featuring advanced machine learning capabilities and intelligent automation. Built with modern web technologies to deliver seamless user experience and robust performance.

Developed for BUET CSE Fest 2023, a comprehensive vaccine registration and management system. Features user registration, appointment scheduling, vaccine inventory tracking, and administrative dashboards for efficient vaccine distribution management.

Built for garbage management system digitalisation of Dhaka city. Implemented features for waste collection scheduling, real-time tracking, and data analytics dashboards.

Designed for students in Bangladesh, integrating multiple AI and virtual learning solutions. Implemented RAG system-based question generation and topic understanding using an AI tutor.

Developed a comprehensive healthcare platform for Bangladesh that integrates telemedicine, online diagnosis appointments, medicine ordering, and ambulance services.