Something created to test the integration of TFJS with React Native and building a classifier with the help of Mobilenet.

IHC based classifier to predict major subtypes of bladder cancer:

OCR-ImageProcessing

Upload an image to find your Bollywood celebrity look-alike

generating mnist-like handwritten numbers using ML methods!

Old ML Project - Create adversarial images to fool a MNIST classifier using TensorFlow.

XGBoost classifiers to predict variant classification conflict

Natural Language Processing with Disaster Tweets

Roman numeral classifier, used to classify hand written roman numerals from 1 to 10 in the form of 28 x 28 pictures. [Ai202L]

Customer_suppot_agent having ability to first classifiy the Query and having two agents that response as per classifer . Agent_FAQ -> Giving Customers Relvant info to Query also have WebSeearch tool for spasific product mention in ship for having more info and physical properties of that product online . 2. out of domian -> respectfully Polite Decl

A deep learning image classifier that probabilistically determines whether a mole is a Melanoma, Nevus or Seborrheic Keratosis.

This project analyzes real feedback from moms about how they use (or avoid) AI in their day-to-day lives. The goal is to uncover patterns in behavior, identify key needs and frustrations, and build personas that represent the major types of AI users within the mom community.

Solo-built conversational AI framework: 143K LoC, 3,250+ tests, 5-tier memory system with cross-encoder reranking, FAISS-indexed Wikipedia (7.2M articles), 8-stage knowledge synthesis pipeline, and full RAG auditability via inline memory citations.