Top Finds

Real-time LLM hallucination guardrail — NLI + RAG fact-checking with token-level streaming halt. Drop-in for any LLM backend.

Linguistic tool (made by a linguist, for linguists) that scraps corpora, automatically cleans it up, and generates n-grams.

🔍 Segment surgical tools using a novel contrastive learning approach in this semi-supervised medical image segmentation network.

AI-powered diabetic retinopathy detection web app using EfficientNet with Grad-CAM visual explanations for interpretable medical diagnosis, developed for research and educational purposes.

🎬 Capture key moments from YouTube videos using heatmaps and create vertical clips for Shorts/Reels/TikTok, complete with AI subtitles.

📊 Forecast time-series data using LSTM models in PyTorch; generate, train, and visualize predictions with key metrics for accurate insights.

This repository contains solutions for Kaggle's Playground Series prediction competitions. It showcases a structured and analytical approach to machine learning, covering both regression and classification tasks.

🖼️ Generate images quickly on Mac Silicon with a lightweight AI model. Enjoy fast processing, local execution, and easy customization without cloud costs.

A Digital Vibe Checker for YouTube, developed by Ananya. This Python tool automates comment extraction via Google API and performs sentiment analysis using NLTK VADER. Generates a visual dashboard (Pie, Bar, Scatter, Time Series) to track community mood vs. engagement. Includes secure .env handling pagination for large datasets and CSV data export.

🎤 Fine-tune the Llasa TTS model with GRPO using Hugging Face tools to enhance performance and evaluate rewards with Whisper ASR and WER metrics.

🎶 Compose high-fidelity music effortlessly by transforming natural language commands into orchestral compositions using AI technology.

🌟 Generate images effortlessly with ComfyUI using the OpenRouter API, supporting multiple inputs and flexible seed control for varied outputs.

🎨 Generate stunning AI images from simple text descriptions with Z-Image Skill for Claude Code. Transform your ideas into visuals effortlessly.

Forensic-grade OCR platform for high-volume document processing. PaddleOCR with Tesseract fallback, language-aware routing, distributed queue processing, and evidentiary integrity preservation.

AI-powered plastic object detection system using YOLO and Streamlit that identifies plastic waste from images and real-time webcam feeds to support smart recycling and environmental sustainability.

🎥 Generate AI-driven videos with Seedance 2.0, offering precise physics, lip-sync, and prompt accuracy for seamless content creation.

Multi-prompt AI text humanizer to reduce detection scores on GPTZero and ZeroGPT

This is personal portfolio

Meta-Agentic Ai Orchestration Platform For Developers and Data Scientists Developing Full-Cycle(E2E) AI Projects

🤖 Explore AI development tools used by companies in Japan, tracking implementations to enhance productivity and innovation in the tech landscape.

A professional-grade, dual-engine algorithmic hedge fund for Alpaca Markets. Features 24/7 continuous trading, ML-driven sentiment analysis (Genetic Algorithms), Numba-optimized backtesting, and a proprietary "Ratchet" risk management system.

🌾 Classify crops using Graph Convolutional Networks for accurate pixel-level analysis, enhancing agricultural insights with cutting-edge machine learning techniques.

Personal news intelligence — aggregate AI content from curated sources, score with LLM, serve your own dashboard

🚀 Optimize FlashVSR to run on low VRAM devices, enhancing video upscaling while minimizing artifacts for smoother playback and better quality.

GNN-based Structural Health Monitoring for CFRP/Al-Honeycomb rocket fairing — debonding defect detection using Graph Neural Networks with FEM-generated datasets