Microscopy is central to biology and neuroscience research; however, manually analyzing thousands of cell images is time-consuming, inconsistent, and prone to human error. Our project, Bio AI: Smart Microscopy Assistant, addresses this challenge by leveraging state-of-the-art deep learning models such as Cellpose and StarDist to automatically segment, count, and analyze cells, nuclei, or organelles from uploaded images. The platform provides robust features, including morphology metrics (size, shape, intensity, eccentricity), batch processing for large datasets, and export options to CSV or Excel for downstream analysis. Deployed via Hugging Face Spaces with a simple Gradio/Streamlit interface, Bio AI ensures accessibility for researchers without programming expertise. By reducing hours of manual work to minutes, the assistant improves reproducibility, scalability, and accessibility of microscopy analysis. Future extensions include disease detection, cancer diagnostics, and drug testing automation, making this tool a high-impact solution for labs, hospitals, and universities.
21 Sep 2025
Our project, Telemedicine_BOT, is a web-based chatbot designed to help users quickly find nearby hospitals, check the availability of doctors and emergency beds, and get health-related guidance. It is especially useful in rural areas where people often don’t have quick access to hospitals or doctor contact information during emergencies. The chatbot is powered by a smart backend using FastAPI, LangChain, Chainlit, and vector search with FAISS, making it responsive and intelligent in understanding user needs. Users just need a phone or computer with internet access to use it. This solution is designed by a team of software engineers working together to build a meaningful healthtech product. It not only saves critical time during emergencies but also reduces costs by avoiding unnecessary travel and wait times. In the future, we plan to add features like live video calls with doctors, image-based disease identification, and offline SMS alerts for low-internet areas. The project has strong real-world potential, especially in countries like Pakistan, where many communities still lack fast access to medical care. Our goal is to build a tool that can help save lives, improve healthcare reach, and create a reliable digital bridge between patients and hospitals.
1 May 2025
The AI-Powered Medical Query Assistant is designed to provide users with quick, accurate answers about medicines, their uses, and potential side effects. By leveraging a vast dataset from PubChem, the app uses advanced AI techniques like sentence embeddings and FAISS indexing to process user queries efficiently. This tool benefits patients, caregivers, medical professionals, and researchers by offering easy access to critical drug-related information. It helps users make informed decisions about medications, avoid harmful side effects, and discover alternative treatments. With its user-friendly interface, this app ensures medical information is readily available for anyone seeking reliable health insights.
26 Jan 2025
marketXminds AI-powered hybrid CRM and POS system enhances business decision-making, optimizes marketing, and improves customer retention with real-time insights. It integrates AI features like product recommendations, personalized pricing, customer segmentation, and churn predictions. Leveraging Meta's Llama 3.2 and Llama 3.1 models, the system combines edge and cloud-based AI to handle both low-latency and complex tasks. 2. Key Features a. Hybrid AI -Edge AI : Lightweight Llama 3.2 models (1B/3B) perform real-time tasks -Cloud AI, Complex tasks, such as large-scale data analysis and pricing optimization, are processed on cloud servers using Llama 3.2 models (11B/90B). b. Llama 3.2 Integration - CRM & POS Insights: Llama 3.2 models predict churn, recommend personalized pricing, and generate strategic business insights. - Real-time Inference: Ensures fast response on edge devices without relying heavily on network connectivity. c. CRM Features : Customer Segmentation & Recommendations, Dynamic Pricing & Promotions. - Customer Retention & Churn Prediction - Automated Campaign Management: Supports AI-driven marketing campaigns for better targeting and results. d. POS Integration - Personalized POS Flow: Offers personalized product recommendations, listings, and dynamic pricing . - Sustainability Incentives: Tracks eco-friendly purchases and rewards customers, while monitoring carbon footprints. e. Advanced Analytics & KPI Recommendations - Business Strategies: Llama 3.2 models analyze KPIs to generate actionable business strategies. - Dashboard: Visually displays KPIs and recommendations. f. Sustainability Insights - Tracking & Incentives: Monitors product carbon footprints and provides suggestions for reducing environmental impact. g. Sentiment Analysis & Follow-ups - Real-time Sentiment: Analyzes customer feedback for improved strategies. - Automated Follow-ups: Sends follow-up actions based on sentiment results.
20 Oct 2024
Flag Globe is an interactive platform designed to make learning about countries engaging and accessible. By allowing users to input either a country name or its corresponding flag emoji, the application instantly provides key information about that country, including capital, population, and a brief historical overview. Utilizing APIs like REST Countries and Wikipedia, Flag Globe also offers an innovative travel feature, which generates travel guidance from the user’s country of origin to the destination country. The goal is to make geography and history education fun, user-friendly, and highly interactive, especially for students and travellers.
13 Oct 2024
.png&w=256&q=75)
.png&w=640&q=75)
.png&w=640&q=75)
