BTC Market Regime Lab

Educational engineering project for market-data processing, probabilistic regime analysis and backtesting applied to BTC.

The goal is to demonstrate a complete technical pipeline—from OHLCV data ingestion and quantitative experimentation to an API and monitoring dashboard. It is not a promise of profitability or financial advice.

This repository is a public, sanitised portfolio version. It does not contain complete datasets, trained models, production parameters, internal Optuna results or real trading signals.

What it demonstrates

• Backtesting workflows with Python, pandas and numpy.
• Probabilistic / Markov-based market-regime experiments.
• Separation between research, inference and visualisation.
• FastAPI backend exposing system state, metrics and operations.
• React, Vite and TypeScript monitoring dashboard.
• PostgreSQL persistence and Docker Compose deployment.

Public architecture

Entorno/
  common/              Shared utilities
  scripts/             Educational data and backtesting scripts
  requirements.txt     Python research dependencies

platform/
  api/                 FastAPI backend
  web/                 React / Vite / TypeScript frontend
  db/                  PostgreSQL initialisation
  docker/              Docker Compose configuration
  docs/                Execution notes and roadmap

Run the platform

From platform/docker:

cd platform/docker
cp ../api/.env.example ../api/.env
docker compose up --build

• Web: http://localhost:8080
• API: http://localhost:8000
• Swagger: http://localhost:8000/docs

For separate frontend and backend development, see platform/docs/EJECUCION.md.

Data

The scripts expect OHLCV data with columns such as:

timestamp,open,high,low,close,volume,source

Real datasets are not published in this repository. For local experiments, use public data from Binance or BitMEX, or provide a small sample CSV under Entorno/data/.

Project status

This started as an early algorithmic research project and evolved into a complete monitoring platform. The public version focuses on architecture, data flow and technical experimentation while keeping private research material out of the repository.

Disclaimer

This project is for education and portfolio demonstration only. It is not financial advice.