Actual Reality Insights

Applied AI research

Peer-reviewed and preprint work spanning applied AI, manufacturing, decision intelligence, and advanced materials.

Analysts refining business intelligence visuals into a clear executive decision dashboard.
Research

Beyond Visualization: Building Decision Intelligence Through Iterative Dashboard Refinement

A four-iteration Power BI study shows how exploratory charts can become a decision-support system for diagnosing profitability, margin, discount, shipping, and market performance.

Three language-model evaluators scoring identical content differently under changing identity cues.
Research

Quantifying Label-Induced Bias in Large Language Model Self- and Cross-Evaluations

Experiments with ChatGPT, Gemini, and Claude show that authorship labels alone can materially change model judgments, supporting blind and diverse multi-model evaluation.

Machine-vision inspection identifying defects in glass with synthetic training examples.
Research

Enhancing Glass Defect Detection with Diffusion Models: Addressing Imbalanced Datasets in Manufacturing Quality Control

Synthetic defect images from a diffusion model improve rare-defect recall across three vision architectures while preserving strong precision for automated glass quality inspection.

An AI agent forming forward and backward plans through a warehouse puzzle.
Research

Interpreting Emergent Planning in Model-Free Reinforcement Learning

Concept-based analysis of a Sokoban agent finds internal representations that predict long-term action effects, causally guide behavior, and function like plans.

Iron-based shape-memory alloy recovering its engineered form under controlled heat.
Research

Manufacturing, Processing, Applications, and Advancements of Fe-Based Shape Memory Alloys

A review of iron-based shape-memory alloys covers conventional and additive manufacturing, characterization, modeling, 4D printing, and structural applications.

Machine-learning model connecting alloy composition and processing to shape-memory actuation strain.
Research

Predicting Actuation Strain in Quaternary Shape Memory Alloy NiTiHfX Using Machine Learning

A machine-learning model links composition, manufacturing, heat treatment, stress, and post-processing to recoverable actuation strain in NiTiHfX alloys.

Showing 6 of 6 insights

Have a real process in mind?

Turn operational friction into a practical AI project.

Tell us where work is slow, repetitive, or hard to scale. We’ll help you find the smallest useful place to start.

Talk through your project