#explainable-ai

Bank reconciliation: when confident AI scoring fails

The case for interpretable design over opaque rankings

Why smarter borrower insights and connected lending workflows must be treated as one transformation…

PEMiG Credit Intelligence Brief | July 20, 2026 Credit Intelligence | Causal AI | Lending Operations | Risk Governance | African Lending

Your SHAP Plot Is Not Explaining What You Think

I built a dataset with the answer key baked in. SHAP, LIME, and permutation importance all agreed on the top features, and all three were…

Graph Neural Networks for Cyber Threat Intelligence: A Research Perspective

Cyber threat intelligence (CTI) relies on structure. Threat actors form networks of infrastructure, malware families, reused credentials…

The Ghost Pixels: How AI Predicts Diseases from Scans, the Dark Bias of Medical Data, and the Human…

Introduction: When a Computer Develops a Medical Eye

Building a Production Churn Pipeline on Databricks: MLflow, SHAP, and Drift Detection

An end-to-end customer churn system built the way a real ML team would ship one — not just a model, but the full lifecycle around it.

Show Your Work

Energy regulation is becoming the hardest test Explainable AI will face. Here is why the source matters more than the answer.

Is Attention Really Explanation? A Debate at the Heart of Explainable AI

The Attractive Power of Attention

My AI Model Predicted Failures. But It Couldn’t Tell Me Why.

The First Lesson I Learned From Machine Learning

From Black Boxes to Transparent Decisions: Why Explainable AI Matters and How We Build It

This article explores how explainable AI reveals model behavior beyond accuracy and why understanding model decisions is as important as…

Can Intelligence be Externalized?

In my previous posts, I explored the idea that intelligence may emerge not only from sheer parameter scale within a single model, but from…

Perturbation XAI Techniques- LIME & SHAP

This article is part of a mini-series on explainable artificial intelligence techniques for deep learning algorithms. This article will…

Explainable Artificial Intelligence (XAI)

Explainable Artificial Intelligence (XAI) aims to make machine learning models transparent, interpretable, and understandable to humans. As…

Building Smarter Medical Applications: AI Models vs. Knowledge-Based Systems

Many fields have benefited from the use of artificial intelligence (AI). One such area where AI developments have greatly benefited is…

DXHC: An AI That Adds Clarity to Forecasting — and to the Models You Already Use

A transparent, physically grounded model that helps you understand your data, validate other models, and make forecasts you can actually…

From a Black Box to a Four-Line Equation: Recovering the Exact Formula a 0.99 Model Was Hiding

A model scored 0.99 on a region that looked nonlinear. A simple diagnostic showed it was four straight lines, and recovered the formula.

Three Devices, One Brain: What I Learned Watching an AI Agent Plumb My Note-Taking Life Together

How a “simple” sync project turned into a live demonstration of why explainability matters more in agents than in models

I Sold Cybersecurity AI for Years. Here’s What Buyers Actually Ask.

Explainability isn’t a research problem. It’s the difference between a cool demo and a signed contract.

Inside the Black Box: 6 Impactful Truths About Explainable AI (XAI)

In recent years, Artificial Intelligence has achieved massive performance momentum, moving from the fringe of computer science to the core…

The Faustian Bargain of Artificial Intelligence

Summary: Society trades cognitive autonomy for AI efficiency, altering our psychology, labor markets, urban spaces, and data privacy.

The Right to Explanation is Enforceable: What GDPR Article 22 and EU AI Act Demand from Your AI…

There is a question that keeps surfacing in AI governance conversations, and it is more legal than it is technical.

Building MalTrace: A Behavioral Malware Analysis Pipeline with Explainable AI

How I built a malware analysis pipeline using CAPEv2 sandbox reports, Random Forest classification, SHAP explanations, MITRE ATT&CK…

LIME vs SHAP: Two Different Detectives Explaining the Same Decision

When a machine learning model gives us a prediction, it usually does not explain the reasoning behind that prediction on its own. A model…

What Building My First Research Paper Taught Me About Machine Learning

When I started my M.Sc. in Data Analytics, I thought machine learning was mostly about building models, improving accuracy, and creating…

Your Company Intranet Is Where Knowledge Goes to Die. Here’s the Engine I’d Build to Fix It.

A practical architecture for an enterprise recommendation system — LightGBM, a hybrid batch + real-time pipeline, and a “why this?” button…

Topological Methods in Modern Artificial Intelligence

Persistent Homology · Mapper · Topological Deep Learning · Explainable AI

5 levels of using AI

5 levels of using AI

Why Building AI for Thai Healthcare Is Harder Than You Think

Part 1 of Building Explainable AI for Thai Healthcare

How Does XAI Actually Work? A Look at SHAP and LIME in Cybersecurity

In my previous piece, we established a problem: AI security systems flag threats without explaining why. Analysts are left staring at…

NN-kNN: A Neural Network That Explains Itself with Neighbors

A visual, code-level guide to the classification and regression versions of NN-kNN

Neural Networks Are Pattern Matchers, Not Physicists.

What 76 Papers Reveal About Where Machine Learning Fails in Science

MiniMax M3: Under the hood for Entry Level Developers

MiniMax M3 is a frontier-level AI model released on June 1, 2026. It is notable for being the first open-weight model to simultaneously…

Why Explainability Is the New Baseline for AI in Governance

Explainable AI in compliance and model governance architecture has shifted from a theoretical tech-stack preference to an absolute baseline…

Attention Based Interpretability With Concept Transformer

Introduction

The Next Wave of AI Progress Might Not Come From Bigger Models

A tiny 27-million-parameter model is outperforming systems 1000x its size on certain reasoning tasks. Here’s what that means.

Why LLM Interpretability Is the Missing Layer in Enterprise AI Success

Artificial intelligence is rapidly becoming a foundational technology across industries. Organizations are integrating advanced AI systems…

What Healthcare Analytics Consulting Actually Delivers: Beyond Dashboards And Data Dumps

Every 24 hours, the average 500-bed hospital generates roughly 137 terabytes of data, yet nearly 80% of that information remains…

I Built an AI That Can Explain Why It Almost Crashed a Car

It started as a resume project. It turned into the most interesting thing I’ve coded.

What SHAP Values Reveal About Climate Risk That Traditional Models Hide

Your model predicts defaults well. But can it explain why — and can a regulator trust the explanation?

Who Deserves Credit When a Phone Has 100,000 Patents?

We propose a new framework borrowed from AI explainability to solve one of the messiest problems in intellectual property law

XAI Research

Opening the Black Box (Qara Qutu): Explaining Telecom Churn (Telecom Müştəri İtkisini İzah…

📎 Bu bloqdakı kod və analizlər üçün GitHub repo: Opening the Black Box — Clean

Opening the Black Box: Explaining Telecom Churn (EN)

📎 Full code and analysis for this blog: Opening the Black Box — Clean

Your Model Made a Decision. Can You Explain It?

A practical guide to SHAP and LIME — two tools that turn black-box predictions into something a human can actually understand

Demystifying the Black Box: Why Explainable AI Is the Future of Machine Learning

Would you trust a colleague who makes major business recommendations but refuses to explain their reasoning? Probably not. Yet this is…

Why Your AI Security System Can’t Explain Itself, And Why That’s Dangerous

Beyond Vectors and Formal Logic: What Ancient Indian Epistemology Can Teach AI About Reasoning

From Mīmāṃsā Hermeneutics to Large Language Models

Tıbbi Yapay Zekanın “Clever Hans” Anı: %99 Doğruluk Bizi Nasıl Kandırır?

Yapay zeka modellerinin tıbbi görüntülemede insanüstü başarılar gösterdiğine dair haberler artık sıradanlaştı. “Zatürreyi %99 doğrulukla…

Your AI Logs Everything. It Can’t Explain Anything.

Why event logs are not epistemic records — and what a justification trace looks like instead

I Built Accurate Healthcare AI Models — But I Still Didn’t Trust Them

Most people think building a good AI healthcare model means achieving high accuracy.