🧍♂️ Core-Correct – Live Human Pose Detection with MediaPipe & Streamlit I built Core-Correct, a real-time human pose detection application that uses a webcam to detect and visualize body landmarks instantly. This project focuses on computer vision fundamentals, real-time processing, and building interactive ML-powered web apps. 🚀 Key Features: * 📹 Real-time pose detection using webcam * 🦴 33 body landmarks with skeletal connections * 💻 Interactive Streamlit-based UI * ⚡ Lightweight, fast, and responsive 🛠 Tech Stack: Python · Streamlit · OpenCV · MediaPipe · NumPy 🎥 Working Demo: https://lnkd.in/d7gZnNpf 📂 GitHub: https://lnkd.in/dRpynh_s This project helped me strengthen my understanding of computer vision pipelines, real-time inference, and UI integration for ML apps. Feedback and suggestions are always welcome! #ComputerVision #MediaPipe #OpenCV #Streamlit #Python #MachineLearning #AIProjects #StudentProjects #CoreCorrect

🚀 Built a GenAI Project: AI Study Buddy 🤖📘 Developed a GenAI-powered web app that "explains concepts" in simple words and generates "MCQ quizzes" for practice. 🛠 Tech: Python (Flask), HTML, CSS, Groq API (LLaMA 3.1) 🔗 GitHub: https://lnkd.in/gCAMZU3g #GenerativeAI #Python #Flask #AIProjects #LearningByDoing

Built and deployed a Car Price Predictor using Machine Learning (Linear Regression) and Streamlit. The app predicts used car prices based on key features and follows a full ML pipeline from data cleaning and EDA to model training and deployment. Live app: https://lnkd.in/gFqEieWQ Github Repo: https://lnkd.in/g-tVUpQQ Always learning, building, and improving. #MachineLearning #DataScience #Python #Streamlit

🚀 Here’s today’s LeetCode practice question I worked on 🐍 Longest Balanced Substring — Simple O(n²) Brute Force Approach (Python) Worked on a problem where we need to find the longest substring in which all distinct characters appear with equal frequency. Instead of over-optimizing, I chose a constraint-aware brute force strategy since n ≤ 1000 — making an O(n²) solution practical and efficient. 🔍 Key Idea A substring is balanced when all character frequencies are equal → min(freq.values()) == max(freq.values()) ⚙️ Approach • Try every possible substring using two loops • Maintain a frequency dictionary incrementally • Check balance using min/max frequency • Track the maximum valid length found ⏱️ Complexity Time: O(n²) Space: O(1) (bounded by alphabet size) 💡 Takeaway: When constraints are small, a clean brute force solution can be more reliable and readable than complex optimizations. Always design based on limits — not assumptions. PYTHON code: class Solution(object): def longestBalanced(self, s): res = 0 for i in range(len(s)): freq = {} for j in range(i, len(s)): freq[s[j]] = freq.get(s[j], 0) + 1 if min(freq.values()) == max(freq.values()): res = max(res, j - i + 1) return res #Python #Algorithms #DataStructures #CodingPractice #ProblemSolving #LeetCode #BruteForce #TechLearning

🚀 I just published a new technical article! How to Connect CopilotKit to a Python Backend Using Direct-to-LLM (FastAPI Guide) In this post, I walk through: Why Direct-to-LLM can be a simpler alternative to heavy orchestration (e.g., LangGraph) How CopilotKit’s Remote Backend Endpoint works A real FastAPI + CopilotKit setup using the official Python SDK A practical in-production use case with streaming responses When this architecture actually makes sense in real systems If you're building AI copilots, agentic apps, or integrating LLMs with existing Python backends, this pattern can save you complexity while keeping things production-ready. Would love your thoughts and feedback! 👇 🔗 Link in comments #CopilotKit #FastAPI #DirectToLLM #Python #AIEngineering #AgenticAI

🎙️Built a Text-to-Speech (TTS) Application using Python! Recently, I worked on a mini project where text can be converted into natural speech in real-time. ✨ Key Features: • Convert text to speech instantly • Adjustable voice speed • Supports long and short sentences • Clean and simple UI ⚙️ Tech Stack Used: • Python • gTTS / pyttsx3 • Tkinter / Streamlit (UI) ✨ Learning Outcome: • Understanding how TTS engines process text • Handling audio generation and playback 🎯 What’s Next: • Planning to add multi-language support • Downloadable audio files • Voice selection Open for feedback. Github:- https://lnkd.in/gctMNDJq

🎬 Movie Recommendation System | Machine Learning Project Built a content-based movie recommendation system using Python, Scikit-learn, and Streamlit. 🔹 Engineered text-based features from genres, keywords, cast & plot   🔹 Used CountVectorizer + Cosine Similarity for content-based recommendations   🔹 Preprocessed and merged 5,000 movies from the TMDB dataset   🔹 Deployed an interactive recommendation web app using Streamlit  This project strengthened my understanding of feature engineering, similarity metrics, and end-to-end ML system design. 🔗 GitHub: https://lnkd.in/gBx-maWD #MachineLearning #Python #RecommendationSystem #Streamlit #ScikitLearn #Projects

The year is young and we’re already shipping new features to enhance our app observability product Honeybadger Insights. 🔍 Insights enables faster, more powerful querying with far less data ingest so you can learn more from your errors and logs. We’re pleased to announce: 1. Auto-refresh features for Insights dashboards. 2. The ability to configure, edit, and share Insights dashboards in YAML. 3. Clickable URLs in Insights query results. Explore these updates on the changelog → https://lnkd.in/gk63p87X #Ruby #Python #ElixirLang

Harnessing Python to perform web searches is unlocking new possibilities in data gathering and automation. By combining powerful libraries like DDGS and httpx, we can seamlessly search, fetch, and process web content programmatically. 🔍 Mastering these tools not only streamlines research but also empowers developers to build tailored search solutions that efficiently parse and rank information. The future of intelligent data retrieval lies in integrating smart algorithms with flexible coding frameworks. The key? A blend of creativity, technical skill, and the drive to innovate. Let’s keep pushing the boundaries of what’s possible with Python. 🚀 #PythonDevelopment #WebSearch #DataAutomation #TechInnovation #Coding #AI

Built a Personality Predictor Web App as an ML project. It predicts whether a person is Introvert, Extrovert, or Ambivert using a Logistic Regression algorithm. 🔹 Features: Normal prediction mode Detailed personality prediction Simple UI Built using HTML, CSS, JavaScript + Machine Learning Deployed as a web application 🔹 Tech Stack: Python | Sklearn | HTML | CSS | JavaScript Live Demo: 👉 https://lnkd.in/gqtY278s This project helped me understand the practical side of ML model integration with web development, model deployment, and user interaction design. #MachineLearning #WebDevelopment #LogisticRegression #Python #MLProject