From how much it matters
to how it works
Go beyond a single importance score to examine the functional patterns in how predictions change with the inputs.
EXPLAINABLE AI · STRIDE XAI
Interpreting the
hidden structures
AI learns.
What has AI learned, and how does it make decisions?
CHAESTRO turns the structure of AI decisions and the interactions behind them into explanations people can understand.
From mathematical structure to real-world questions.
STRIDE XAIBeyond importance.
Beyond importance.
Into the structure of decisions.
STRIDE is an explainable AI research framework that decomposes model predictions into functional components, revealing how individual factors act and interact.
From how much it matters
What factors reveal
when they act together
Separate patterns that individual factors alone cannot explain, and analyze how a model combines multiple conditions.
Connecting research
to real-world data
Define the model, data and practical questions together, then explore STRIDE's applicability through a proof of concept (PoC).
RESEARCH PAPER · ARXIV PREPRINTSTRIDE: Subset-Free Functional Decomposition
Read the paper
STRIDE: Subset-Free Functional Decomposition
for XAI in Tabular Settings
Chaeyun Ko · 2025 · Functional decomposition and interaction analysis for tabular models
The paper's findings are limited to the models, datasets and assumptions studied. Industry-specific outcomes are assessed through individual PoCs.
FOUNDERExploring AI through economics,
Exploring AI through economics,
philosophy and mathematics.
Chaeyun Ko
Creator of STRIDE XAI
Ewha Womans UniversityMajors in Economics and Philosophy · Minor in Mathematics
Master's degree in Mathematics
STRIDE XAIExplainable Artificial Intelligence
Research on AI decision structures and interactions
Economics taught me to bring structure to complex phenomena. Philosophy led me to questions of truth, essence and knowledge. Mathematics gave me a language for turning intuition into rigorous structures and proofs. I went on to develop STRIDE XAI through explainable AI research, and founded CHAESTRO to bring that work into technology, products and a company.
Explainable AIMathematical ModelingResearch & Development
Making AI decisions
understandable
Ongoing research and development to interpret the patterns models learn and connect them to practical problems.
One PCT international application
Novelty · Inventive step · Industrial applicability
Favorable opinion on all three criteria
Claims 1–12 · Written Opinion of the International Searching Authority
August 19, 2026 · Filed in the founder's name
This is an opinion at the international search stage, not a granted patent.
PoC · INVESTMENT · RESEARCH COLLABORATION
Tell us what you'd
Tell us what you'd
like to explore.
We welcome companies exploring STRIDE XAI PoCs in finance, healthcare, manufacturing and beyond, as well as investors and research institutions interested in CHAESTRO's research and technology.
Industry PoCsInvestmentResearch & technology partnerships
Send a short introduction with your organization, interests and what you'd like to discuss.