🏀 Build and Deploy an AI NBA Game Prediction App — Step-by-Step (Week 6) In this final episode of the NBA Insights series, we deploy our XGBoost-powered prediction model to Hugging Face Spaces using Streamlit, creating a live web app that predicts NBA game outcomes from user input or CSV uploads. 💡 This video is perfect for anyone looking to: • Deploy machine learning models with Streamlit • Use Hugging Face Spaces for app hosting • Handle deployment issues like Docker image errors and environment variable secrets • Integrate Google Sheets API securely with a deployed app • Share live ML demos without building from scratch 🔥 What You’ll Learn: • Why we switched from Docker SDK to Streamlit on Hugging Face • How to safely store and access API keys & secrets using Hugging Face Secrets • How to upload a .pkl model as a Hugging Face Dataset (to fix common deployment issues) • How to expose a simple yet effective ML UI for manual and batch (CSV) predictions • Real-world lessons from deploying Streamlit apps to the cloud 📊 The app also integrates Tableau dashboards for real-time game insights and was built using: • nba_api, pandas, XGBoost, Streamlit, Hugging Face, gspread, and Tableau 👉 Watch the full project breakdown here: https://www.youtube.com/playlist?list=PLnYNA149f2AUc2HKSZ7HjlWJM4Yr_H_R0 🔮 Try the live app: 👉 https://huggingface.co/spaces/aonat/nba_insights_streamlit Catch up on previous weeks): 📖 https://www.aykutonat.com/post/from-passion-to-prediction-building-the-nba-insights-system-weeks-1-2-recap 📖 https://www.aykutonat.com/post/building-nba-insights-ai-weeks-3-to-5-recap 👉 Follow the project and explore the code: 🔗 GitLab Repository: https://gitlab.com/aykutonat/nba_ml_insights 🌐 Website: www.aykutonat.com 💼 LinkedIn: linkedin.com/in/aykutonat 👍 Like, 💬 comment, and 🔔 subscribe for more weekly episodes! #NBA #nbaapi #pythonproject #datascience #basketballanalytics #NBAInsights #nba2025 #nbaprops #nbapredictions #nbaodds #odds #nbapicks #nbapickstoday #nbapredictionstoday #nbabets #nbaplayoffs #automation #machinelearning #deeplearning #ai #aiautomation #nbapredictions2024 #nbaaiprediction #sportsanalytics #nbadatascience #featureengineeringpython #nbamachinelearning #nbagamepredictionai #xgboostnba #logisticregressionnba #pandasrollingstats #nbaanalytics #datascienceproject #Streamlit #HuggingFace

Can an AI model trained on past NBA data predict the future of the league? In this episode, we put our custom XGBoost-powered NBA game prediction model to the test — using completely unseen 2024–2025 season data. 🔍 We walk through: New feature engineering strategies (win streaks, opponent win rate, rest days & more) Hyperparameter tuning with GridSearchCV Model performance analysis on real future games Calibration curves and evaluation metrics 🚀 Final results? 70% accuracy on unseen games and a well-calibrated probability model — ready to simulate and predict weekly matchups when the 2025–26 season kicks off. 📌 Stay tuned — we’ll be back with live predictions and further model upgrades next season! 👉 Watch the full project breakdown here: https://www.youtube.com/playlist?list=PLnYNA149f2AUc2HKSZ7HjlWJM4Yr_H_R0 Catch up on previous weeks (Data Pipeline + Feature Engineering): 📖 https://www.aykutonat.com/post/from-passion-to-prediction-building-the-nba-insights-system-weeks-1-2-recap 👉 Follow the project and explore the code: 🔗 GitLab Repository: https://gitlab.com/aykutonat/nba_ml_insights 🌐 Website: www.aykutonat.com 💼 LinkedIn: linkedin.com/in/aykutonat 👍 Like, 💬 comment, and 🔔 subscribe for more weekly episodes! #NBA #nbaapi #pythonproject #datascience #basketballanalytics #NBAInsights #nba2025 #nbaprops #nbapredictions #nbaodds #odds #nbapicks #nbapickstoday #nbapredictionstoday #nbabets #nbaplayoffs #automation #machinelearning #deeplearning #ai #aiautomation #nbapredictions2024 #nbaaiprediction #sportsanalytics #nbadatascience #featureengineeringpython #nbamachinelearning #nbagamepredictionai #xgboostnba #logisticregressionnba #pandasrollingstats #nbaanalytics #datascienceproject #nbafeatureengineering #scikitlearnnba #pythonnbaprediction #nbaaimodel #sportsdatascience

In Week 4 of the NBA Insights project, we test our AI model's ability to predict real NBA game outcomes using advanced sports performance analysis techniques. We built a custom prediction pipeline with XGBoost, created an interactive Streamlit app for manual and batch simulation, integrated Google Sheets for seamless export, and embedded Tableau dashboards to visualize team trends and model accuracy. This episode blends machine learning prediction project work with design and analysis of algorithms, and gives you a hands-on look at building an AI agent for data analysis — applied to real sports data. Whether you're exploring eCommerce product research tools, artificial intelligence workshops, or simply want to simulate NBA outcomes with real data, you're in the right place. 👉 Follow the project and explore the code: 🔗 GitLab Repository: https://gitlab.com/aykutonat/nba_ml_insights 🌐 Website: www.aykutonat.com 💼 LinkedIn: linkedin.com/in/aykutonat ✅ Week 4 Highlights: Inference pipeline to predict game outcomes Streamlit app for manual and CSV-based simulation Google Sheets integration for exports Embedded Tableau dashboards: shooting % trends, win/loss predictions, confidence validation Keywords: NBA game predictions, predictive modeling in sports, Streamlit apps, AI-powered analytics, machine learning for basketball, sports data engineering, XGBoost, GitLab AI projects Catch up on previous weeks (Data Pipeline + Feature Engineering): 📖 https://www.aykutonat.com/post/from-passion-to-prediction-building-the-nba-insights-system-weeks-1-2-recap Whether you're into sports analytics, AI, or just curious if machines can predict the NBA — this series is for you. 👍 Like, 💬 comment, and 🔔 subscribe for more weekly episodes! #NBA #nbaapi #pythonproject #datascience #basketballanalytics #NBAInsights #nba2025 #nbaprops #nbapredictions #nbaodds #odds #nbapicks #nbapickstoday #nbapredictionstoday #nbabets #nbaplayoffs #automation #machinelearning #deeplearning #ai #aiautomation #nbapredictions2024 #nbaaiprediction #sportsanalytics #nbadatascience #featureengineeringpython #nbamachinelearning #nbagamepredictionai #xgboostnba #logisticregressionnba #pandasrollingstats #nbaanalytics #datascienceproject #nbafeatureengineering #scikitlearnnba #pythonnbaprediction #nbaaimodel #sportsdatascience

Welcome back to NBA Insights — a weekly AI project where we build an NBA game prediction system from scratch using real data and machine learning! We’re taking everything we've built so far and turning it into our first real predictive engine — using logistic regression and XGBoost to classify game outcomes based on 5-game rolling averages of NBA team stats. This is where data starts making decisions. 📌 In this episode, you’ll learn: ✔️ How we trained our baseline model using logistic regression ✔️ Why accuracy isn’t everything — and how to interpret precision, recall, and F1 scores ✔️ What the confusion matrix tells us about wins vs. losses ✔️ How to use the ROC AUC score to judge model performance ✔️ What worked, what didn’t, and how we’ll improve from here View the full code and model training notebook on GitLab: 🔗 https://gitlab.com/ayk.onat/nba_ml_insights Catch up on previous weeks (Data Pipeline + Feature Engineering): 📖 https://www.aykutonat.com/post/from-passion-to-prediction-building-the-nba-insights-system-weeks-1-2-recap Whether you're into sports analytics, AI, or just curious if machines can predict the NBA — this series is for you. 👍 Like, 💬 comment, and 🔔 subscribe for more weekly episodes! #NBA #nbaapi #pythonproject #datascience #basketballanalytics #NBAInsights #nba2025 #nbaprops #nbapredictions #nbaodds #odds #nbapicks #nbapickstoday #nbapredictionstoday #nbabets #nbaplayoffs #automation #machinelearning #deeplearning #ai #aiautomation #nbapredictions2024 #nbaaiprediction #sportsanalytics #nbadatascience #featureengineeringpython #nbamachinelearning #nbagamepredictionai #xgboostnba #logisticregressionnba #pandasrollingstats #nbaanalytics #datascienceproject #nbafeatureengineering #scikitlearnnba #pythonnbaprediction #nbaaimodel #sportsdatascience

Welcome to Week 2 of the NBA Insights Project — where we dive deep into feature engineering for AI-powered NBA predictions using the 2024 regular season data. In this episode, you'll learn how to: * Create rolling averages for team stats (PTS, AST, REB, FG%) * Add opponent performance metrics * Engineer contextual features like home-court advantage * Prepare clean and predictive datasets for modeling This is part of a complete machine learning pipeline for NBA game prediction, ideal for data science learners, sports analysts, and AI enthusiasts. 📈 Code used: Python, Pandas, Scikit-Learn 🔮 Next episode: Modeling NBA games with Logistic Regression and XGBoost 🚀 Full project & resources: www.aykutonat.com 📌 Subscribe to follow the full AI series #NBA #nbaapi #pythonproject #datascience #basketballanalytics #NBAInsights #nba2025 #nbaprops #nbapredictions #nbaodds #odds #nbapicks #nbapickstoday #nbapredictionstoday #nbabets #nbaplayoffs #automation #machinelearning #deeplearning #ai #aiautomation #nbapredictions2024 #nbaaiprediction #sportsanalytics #nbadatascience #featureengineeringpython #nbamachinelearning #nbagamepredictionai #xgboostnba #logisticregressionnba #pandasrollingstats #nbaanalytics #datascienceproject #nbafeatureengineering #scikitlearnnba #pythonnbaprediction #nbaaimodel #sportsdatascience

🏀 Welcome to Week 1 of the NBA Insights project! In this episode, we pull real NBA game data from the 2023–24 season using Python and the nba_api. We explore early trends, dive into rolling win rates, and spot momentum swings in teams like the Lakers. Tools used: - nba_api - pandas - Jupyter Notebook What we cover: ✔️ Fetching NBA game logs ✔️ Saving to CSV ✔️ Exploring win/loss trends ✔️ Rolling average visualizations ✔️ How to cleanly analyze without modifying raw data 📊 Next week: Feature engineering — we’ll create smarter stats and prep this data for machine learning. Built by a data pro and basketball fan. #NBA #nbaapi #pythonproject #datascience #basketballanalytics #NBAInsights #nba2025 #nbaprops #nbapredictions #nbaodds #odds #nbapicks #nbapickstoday #nbapredictionstoday #nbabets #nbaplayoffs #automation #machinelearning #deeplearning #ai #aiautomation

Tired of digging through spreadsheets to answer simple Amazon product questions? In this video, I’ll show you how to build an AI-powered catalog chatbot that instantly responds to product queries using your structured Amazon catalog data. Whether you’re an Amazon seller, e-commerce manager, or catalog specialist — this chatbot can transform how you handle product information. What You’ll Learn: *** Why catalog chatbots solve a real pain point in Amazon operations *** How to generate product embeddings using sentence-transformers *** How to build a clean, interactive UI with Gradio (no front-end coding!) *** How to enable voice queries and match natural language with product attributes Future integrations: Shopify, Slack, multilingual, and built-in checkout Have questions or want to collaborate? 🔗 Find me on LinkedIn: linkedin.com/in/aykutonat/ #ai #artificialintelligence #future #futuretools #machinelearning #deeplearning #aitools #gradio #shopify #amazon #amazonproduct #chatbots #voicechat #bots #ui #uiux #automation #automationtools #automationwithpython #aiautomation #bestaitools #bestaitools2025 #freeaitools #topaitools #aiforwork #aiformarketing #aicontentcreation #aiforcreators

Discover how to instantly turn product images into high-converting Amazon listings using AI. In this demo, I showcase a bulk image SEO generator that auto-generates product titles and bullet points for e-commerce sellers. Perfect for catalog managers, Amazon sellers, and Shopify merchants looking to automate tedious content work. Full Description: This tool uses computer vision and GPT-4 to: Analyze product images Generate a caption Create an Amazon-style product title Write 5 SEO-optimized bullet points Export results to CSV for bulk upload ▶ Built with Gradio, Python, BLIP & OpenAI 🔗 Try it out or hire me: [Insert link to your site or Upwork profile] 🌐 More AI tools & demos: https://aykutonat.com #ai #artificialintelligence #future #futuretools #machinelearning #deeplearning #aitools #AmazonSEO #EcommerceAI #GPT4 #ProductListingAutomation #AIforRetail #seo #gradio #python #blip #openai #automation #automationtools #automationwithpython #aiautomation #bestaitools #bestaitools2025 #freeaitools #topaitools #aiforwork #aiformarketing #aicontentcreation #aiforcreators