Built a Streamlit app for a coffee shop in 100 Days Challenge

Excited to share my latest Machine Learning project: a Soccer Tactical Map Generator! ⚽️🗺️ This tool uses Computer Vision to analyze a football match image, detect keypoints on the field (like penalty corners) and the referee, and then generate a 2D tactical map showing the players' positions. The core project was built to handle both video and static images, and as you can see in the demo, the deployed web app works perfectly with picture uploads. A special thanks to our supervising professor, Bram Heyns, for his guidance throughout this project. Tech Stack: Core: Python Deployment: Flask & Docker Want to check it out? You can see all the code and details on my GitHub: https://lnkd.in/gtGzu5Xg Or, you can run the app directly using Docker: Pull the image: docker pull jfgm299/tactical-map-app-final Run the container: docker run -p 5001:5001 jfgm299/tactical-map-app-final Open your browser to http://localhost:5001 This was a fantastic challenge, blending my interests in computer vision and giving me a real vision on how to work in bigger projects. Check out the quick demo video to see the clean UI and how it works! #MachineLearning #ComputerVision #Python #Flask #Docker #SportsAnalytics #Soccer #Homography #AI #TechProject

Thrilled to share my new project — Visual Product Matcher It’s a Streamlit web app that uses CLIP (ViT-B/32) embeddings to find visually similar products! 🛍️ You can upload an image or paste an image URL, and the app searches through a catalog of 1000+ products to show the most visually similar matches — complete with thumbnails, metadata, and similarity scores. 💡 Highlights: Image upload & live preview Visual similarity search using CLIP Adjustable filters (Top-K & similarity threshold) Cached responses for speed Clean, mobile-responsive Streamlit interface Tech Stack: Python | Streamlit | Sentence-Transformers | CLIP | JSON API This project combines AI + Computer Vision + Interactive Dashboards — all running seamlessly on free tiers! #AI #MachineLearning #DeepLearning #ComputerVision #Streamlit #CLIP #Python #DataScience #Project #Innovation

I'm thrilled to share a project I've been working on: an end-to-end Customer Churn Predictor! 🚀 It's one thing to learn the theory of machine learning, but building a real, interactive application from scratch is a whole different challenge. I wanted to take a dataset and bring it to life. Here’s what I did: 🔹 The Goal: Predict if a bank customer would churn using a dataset from Kaggle. 🔹 The Brains: I trained a Random Forest model that achieved 86% accuracy in identifying at-risk customers. 🔹 The Interface: I built a clean, professional web app, allowing anyone to get instant predictions. Check out the video to see it in action! 👇 This project was a fantastic learning experience in data preprocessing, model evaluation (balancing precision vs. recall), and front-end development. #MachineLearning #DataScience #Python #Streamlit #PortfolioProject #DataAnalytics #PredictiveAnalytics #Project

🎬 Excited to share my Advanced Movie Comparison & Recommendation App! MovieFlix!!! Built with Streamlit, this tool offers: ✅ Personalized movie recommendations using content-based filtering ✅ Side-by-side movie comparison feature ✅ AI-powered chatbot for interactive movie queries ✅ Movie posters & detailed information via OMDB API ✅ Fetching movie trailer in the live app using YouTube API ✅ Intuitive search functionality ✅ Real-time similarity analysis This project helped me apply data science and machine learning concepts in a practical, user-friendly interface. Special thanks to Nitesh Dhar Badgayan (PhD) Sir, from KPMG, for the invaluable guidance and to Globsyn Business School for fostering an environment of innovation and learning. 🔗 Try it here: https://lnkd.in/g3_JpMPX #DataScience #MachineLearning #RecommendationSystem #Streamlit #Python #MovieRecommendation #AIChatbot

🚀 Introducing Live Preview in Mercury! Building web apps from Jupyter Notebooks just became even smoother. With Live Preview, you can now create your app and see the results instantly — side by side. ✅ Edit your notebook ✅ Adjust widgets and settings ✅ See the live app update in real time ✅ No refreshing, no re-running, no friction This feature turns Mercury into a true app-building workspace, perfect for: data analysts, educators, domain experts, ML engineers. Your workflow stays simple. Your app stays visible. Your creativity stays in flow. Try it out in Mercury and experience how effortless notebook-to-app creation can be. 🚀 https://RunMercury.com 👉 https://lnkd.in/eV7UuMSS #Mercury #MLJAR #AutoML #Jupyter #WebApps #DataScience #Python #OpenSource #Productivity #Innovation #AI #AIAgent

🌸 Proud to share my very first Machine Learning Web App — Iris Flower Classification! I developed this project using Python and Streamlit in PyCharm, integrating both model training and web deployment. The app predicts the species of an Iris flower (Setosa, Versicolor, or Virginica) based on four input parameters — sepal length, sepal width, petal length, and petal width. 💡 Project Workflow & Highlights: 🔹 Trained a Random Forest Classifier on the Iris dataset using Scikit-learn 🔹 Split data into train/test sets for accurate evaluation 🔹 Saved the trained model using Pickle for later use 🔹 Deployed the model with Streamlit to create an interactive web app 🔹 Designed a clean UI that provides real-time predictions This being my first end-to-end Machine Learning app, it was a great hands-on experience that strengthened my understanding of the complete ML pipeline — from data preprocessing to model deployment. 💻 GitHub Repository: https://lnkd.in/gSZAVx5K 📸 (Screenshot of the web interface below) #MachineLearning #Python #Streamlit #ScikitLearn #PyCharm #DataScience #AI #RandomForest #ModelDeployment #GitHubProjects #FirstProject #LearningJourney #MLProjects

🚀 Learning by Building: My Weather App Journey 🌦️ I recently built a Weather App — and while I haven’t invented anything groundbreaking, this project turned out to be an incredible learning experience! During development, I explored tools that were completely new to me: Streamlit – I learned how to create a fully interactive and user-friendly web interface, making the app accessible and visually appealing. Plotly – I dived into interactive data visualization, creating dynamic charts to represent weather patterns clearly and intuitively. Alongside these, I also leveraged: Requests – To fetch real-time weather data from the OpenWeatherMap API. Pandas – For organizing, cleaning, and processing data efficiently. Datetime – To handle timestamps and display accurate date & time information. Python-dotenv – To securely manage my API key and environment variables. Through this project, I not only reinforced my Python skills but also gained hands-on experience with APIs, data handling, and interactive UI/visualizations — all in a single application. It’s amazing how a small project can teach so many concepts at once. This journey has expanded my toolkit and boosted my confidence to take on more real-world projects! 💡 #Python #Streamlit #Plotly #DataVisualization #APIs #LearningByDoing #ProjectShowcase #TechJourney Github URL: https://lnkd.in/gNaNWbgY

🌟 Exciting Project Update! 🌟 I’m thrilled to share the launch of AI-GARBAGE-DETECTOR, a Python-powered solution aimed at automatically identifying waste in images using computer vision. 🔍 What it does: Utilises YOLOv8 (via the file yolov8n.pt) to detect and classify garbage items in imagery. Built in Python with a simple app.py interface. GitHub 🚀 Why it matters: Helps in waste-management and clean-city initiatives Enables quicker identification of litter hotspots or bin-overflows Provides a practical “real-world” AI use case merging my data analysis & Python programming interests 💡 What I learned: Integrating object detection models into runnable apps Handling Python dependencies & packaging (see requirements.txt) GitHub Thinking end-to-end: from model files to user interface 🔧 What’s next: Expanding the dataset for more garbage types Improving detection accuracy & speed Deploying as a micro-service / mobile app for real-time usage Collaborating with municipalities or environmental organisations 📣 If you’re interested in AI + sustainability, I’d love to connect! Feel free to check out the repository here → [GitHub link] and share feedback, ideas or potential collaborations. #ArtificialIntelligence #ComputerVision #Python #Sustainability #DataScience #MachineLearning

Many GTM teams lose momentum between analysis and execution. But it doesn't have to be that way. That’s where AI tools like Dataiku, Python, Visual Studio Code, Lovable, and Lindy come in. Together, they turn GTM ideas into live workflows, mini-apps, and automations — built from natural language, not tickets. Check out our short guide below on how we connect these tools to accelerate AI-driven GTM execution. #AIforGTM #GoToMarket #MarketingInnovation #LeanGTM #AIagents