Model explainability that works seamlessly with 🤗 transformers. Explain your transformers model in just 2 lines of code.

Diffusers-Interpret 🤗🧨🕵️‍♀️: Model explainability for 🤗 Diffusers. Get explanations for your generated images.

A curated list of awesome open source tools and commercial products for monitoring data quality, monitoring model performance, and profiling data 🚀

A Python client to interact with Arize API

code for studying OpenAI's CLIP explainability

CrysXPP: An Explainable Property Predictor for Crystalline Materials (NPJ Computational Materials - 2022)

A minimal, reproducible explainable-AI demo using SHAP values on tabular data. Trains RandomForest or LogisticRegression models, computes global and local feature importances, and visualizes results through summary and dependence plots, all in under 100 lines of Python.

The official Python library for Openlayer, the Continuous Model Improvement Platform for AI. 📈

Explainability toolkit for retrieval models. Explain prediction of vector search models (embeddings similarity models, siamese encoders, bi-encoders, dense retrieval models). Debug your vector search models for RAG or agentic AI system.

The Fraud Detection project aims to improve identification of fraudulent activities in e-commerce and banking by developing advanced machine learning models that analyze transaction data, employ feature engineering, and implement real-time monitoring for high accuracy fraud detection.

A comprehensive archive of completed Kaggle Data Science, Machine Learning, and Data Manipulation courses, documenting structured Python exercises and certifications.

A Java client to interact with Arize API

Capture fundamentals around ethics of AI, responsible AI from principle, process, standards, guidelines, ecosystem, regulation/risk standpoint.

Explaining Trees (LightGBM) with FastTreeShap (Shapley) and What if tool

Example projects for Arthur Model Monitoring Platform

Study on the performance of pre-trained models (VGG16, EfficientNetb0, ResNet50, ViT16) with weight fine tuning, as well as classical ML algorithms (Naive Bayes, Logistic Regression, Random Forest) on a dataset of 6.806 fungi microscopy Images utilizing Pytorch.

A proof-of-concept for the implementation of an early fault detection system in oil wells, designed to enhance operational efficiency and reduce costs.

🏦 Build a complete data engineering workflow for a banking system, showcasing ETL processes, data transformations, and an interactive financial dashboard.

A reusable codebase for fast data science and machine learning experimentation, integrating various open-source tools to support automatic EDA, ML models experimentation and tracking, model inference, model explainability, bias, and data drift analysis.

Machine Learning Individual Project - November 23, 2021

This project is a machine learning competition hosted on Kaggle platform, focused on forecasting Walmart's monthly and quarterly sales. We tasked with developing advanced predictive models to accurately predict Walmart's sales, taking into account various factors such as historical sales data, macroeconomic indicators, and local market conditions.

Examples of problem-solving in Information Retrieval, RAG, and Agentic Systems

Production-ready ML model predicting DoorDash delivery times.

PyTorch implementation of influence functions: ICML 2017 method, TracIn (NeurIPS 2020) and EmpiricalIF (NeurIPS 2022). Estimate how each training sample affects model predictions without retraining.

Portfolio of real-world ML projects demonstrating ranking & recommendation systems, engagement prediction, fairness, and explainability, engineered end-to-end with scalable, production-ready design principles.

Developed an efficient system using Regression Techniques to empower retailers with profitable insights & maintain a competitive edge in the dynamic retail industry.

Deep learning–based pneumonia detection in chest X-rays with clinical threshold optimization and Grad-CAM explainability.

ai powered loan approval prediction system built using machine learning and streamlit. the project analyzes applicant financial data to predict loan approval probability, generate risk scores, provide model insights, and support data driven credit decision making through an interactive analytics dashboard.

This project provides a performance evaluation of credit card default prediction. Thus different models are used to test the variable in predicting the credit default and we found Random Forest Classifier performs the best with a recall of 0.95 on the test set.

An application of the WhizML codebase for an analysis of cardiovascular disease risk.