Hi, I'm Ivan Gorban.
Staff Data Scientist working on machine learning, causal inference, experimentation, marketplace systems, and applied AI.
I build and study systems that help products make better decisions: pricing, churn prevention, marketplace optimization, experiment design, and AI-assisted workflows. I also write about causal inference, automation, research practice, and how AI changes learning and knowledge work.
Areas I focus on
-
Machine learning for products
Applied machine learning systems for real product and marketplace decisions: pricing, churn prevention, lifecycle modeling, customer and supply behavior. Models that do not just predict outcomes, but help teams decide what to do next.
-
Causal inference & experimentation
Experiment design, causal measurement, uplift estimation, switchback tests, and practical methods for separating correlation from intervention effects.
-
Applied AI & LLM systems
Practical applications of AI and LLMs in workflows, tooling, education, and knowledge work — agentic coding, AI-assisted analysis, tutoring systems.
-
Research & writing
Causal inference, experimentation, AI, automation, data science practice, and the economic and educational consequences of new technology.
Featured work
All work →-
Quoin ↗
in progressA developer workflow library for AI-assisted coding. Structured planning, implementation, and review skills — built so developers stay in control while leveraging Claude.
-
AI teacher for a kid
in progressA personal project exploring how AI tutors can help kids learn — built on the OpenClaw platform — specifically, whether an LLM-powered system can teach a child effectively while keeping human judgment in the loop. The project investigates how to structure conversations, handle misconceptions, and build genuine understanding rather than surface-level pattern matching. Currently in early-stage experimentation and not yet shareable publicly.
Latest writing
All writing →-
The ML Trap in Price Elasticity Estimation
MediumHow an ML model can predict purchasing patterns well yet estimate price elasticity badly — and why endogenous pricing leads flexible models to learn spurious correlations and recommend wrong prices.
-
Quoin: From Prompt Engineering to Workflow State for Claude Code
MediumWhy Quoin moved from prompt engineering to a workflow-state model when integrating Claude Code into real codebases.
-
Demystifying Causality: An Introduction in Causal Inference and Applications. Part 5
MediumInstrumental variables in causal inference — using natural experiments like rivers and judicial assignment to identify causal effects.