tensor-house

Why we included this project

For data scientists and ML engineers who need to prove an AI idea works before investing in a full pipeline, TensorHouse collects worked Jupyter notebooks for the problems that keep showing up in enterprise work: pricing, marketing, supply chain, and smart manufacturing. Each notebook pairs a modeling approach with simulated or sample data, so you can test whether a technique fits your problem before wiring up your own data pipeline. The project deliberately mixes deep learning, reinforcement learning, causal inference, and Bayesian methods, many of them built by practitioners or researchers who worked with retailers and manufacturers. It also ships readiness questionnaires for scoping what data and integration work a use case actually demands. That combination makes it a practical starting point for turning a vague business question into a prototype you can show stakeholders.