Prismatic Metrics

🔎Deep Research Agent 👨‍💻 Author: Shailesh Gupta 🔗 GitHub: View Source Code 🧠 What is this project about? This is an agentic Deep Research Assistant built using Gradio , OpenAI Agents SDK , Web Search , Pydantic , and SendGrid . Unlike a simple chatbot that gives one quick answer, this system behaves more like a structured research assistant. It first checks whether the user’s query is clear enough, asks 3 to 4 clarifying questions when necessary, creates a targeted research plan , performs multiple web searches , collects structured evidence, and finally generates a detailed long-form research report . The system also uses an orchestrator-agent architecture , where a top-level manager coordinates specialized agents for clarification, planning, searching, writing, and delivery. In simple terms, this project was built to simulate how a real AI-powered research workflow should operate — step by step, with reasoning, structure, evidence, and synthesis. 📦 Key Features ✅ Clarificatio...

🎬 Project: Movie Recommender System 👨‍💻 Author: Shailesh Gupta 🔗 GitHub: View Source Code 🧠 What is this project about? This is a content-based movie recommendation system that suggests similar movies based on your selected title. It uses movie metadata (genres, tags, cast, crew) to compute similarity scores and recommend relevant titles. The model uses TF-IDF vectorization and cosine similarity to find the top 5 similar movies from a large movie dataset. 📦 Key Features: Combines features like overview, cast, genres, and keywords into a single vector Similarity computed using cosine_similarity from scikit-learn Interactive Streamlit frontend with easy-to-use dropdown for movie selection Top 5 movie recommendations displayed with titles and (where available) posters ⚠️ Deployment Note: This app could not be hosted on Streamlit Cloud due to GitHub file size limitations. Some files like tmdb_5000_movies.csv and pickled models are large and ...

📌 Project: Stock Price Predictor 👨‍💻 Author: Shailesh Gupta 🔗 GitHub: View Source Code 📈 Objective: Develop a predictive model that forecasts future stock prices using historical data — combining time series insights with deep learning techniques. 📊 Use Case: Predicting next-day or next-week stock trends can give traders, analysts, and retail investors a valuable edge. This model is focused on learning **price patterns** from past data and providing **next-value estimations**. 🔍 Dataset Overview: Used a historical CSV of daily stock prices Preprocessed with MinMax scaling Split into training/testing sequences using time steps 🧠 Model Highlights: LSTM-based model trained on sequences of past prices Captures time dependencies and patterns in sequential data Plots predicted vs actual values to visualize model performance 🛠️ Tools Used: Python, Pandas, NumPy Keras & TensorFlow (LST...

📌 Project: Campus Placement Predictor 👨‍💻 Author: Shailesh Gupta 🔗 GitHub: View Source Code 💡 Q: What problem does this project solve? Many students go through their campus years wondering: "Will I get placed?" This project aims to answer that question — by building a machine learning model that predicts the likelihood of campus placement based on a student’s academic profile and background. 🛠️ Q: What kind of data is used? The model uses structured data containing: Academic scores (SSC, HSC, degree, MBA) Work experience Specialization Gender, streams, and placement status It’s a real-world classification task where the output is binary: Placed ✅ or Not Placed ❌ 🔍 Q: How does the model work? After cleaning and preprocessing the data, I trained models using: Logistic Regression Decision Trees Evaluation metrics like accuracy, confusion matrix, and ROC-AUC were used to validate performance. 💬 Q: Can I try it live? Ab...

📌 Project: Udemy Course Recommendation System 👨‍💻 Author: Shailesh Gupta 🔗 GitHub: View Source Code 🎯 What’s this about? Ever spent more time choosing a course than actually taking one? This project is my take on solving that. It’s a content-based recommendation system that filters Udemy courses by matching your interest with course descriptions using TF-IDF and cosine similarity . "The right course, for the right learner, at the right time." 📘 Use Case: Let’s say you type “machine learning beginner” . The system ranks similar Udemy courses by calculating how closely the description matches your query — like a search engine, but smarter ✨. 🧠 Key Concepts Applied: TF-IDF Vectorization for text embedding Cosine Similarity for content matching Pandas & NumPy for data wrangling Streamlit for deploying the app 🚀 Try it out: 👉 Launch the Recommendation App Here’s what it looks like: 🔍 What You’ll Find in the Repo...

Project: Hamlet Next Word Prediction using LSTM Author: Shailesh Gupta GitHub: View Source Code 📌 Overview This deep learning project uses an LSTM (Long Short-Term Memory) network to perform next word prediction on a text corpus derived from Shakespeare's Hamlet . The model learns word sequences and generates text one word at a time, simulating the creative structure of natural language. 🎯 Problem Statement Can a neural network trained on classical literature predict the next word in a sentence with contextual accuracy? This project explores the ability of LSTM-based models to understand and generate language sequences from complex, stylized text. 🔍 Key Features Text generation using LSTM layers in Keras Custom tokenizer, sequence creation, and word embedding Model trained on over 12,000 words from Shakespeare’s Hamlet User input allows dynamic next-word prediction Deployed as an interactive Streamlit app 🛠️ Tools & Tech Stack Python, T...