Bottie — build the trading stack

Welcome to the course book for Bottie: a chapter-based path through a real FastAPI backend, execution engine, ML hooks, optional quantum-inspired portfolio code, and sentiment integration—as documented here. This repository holds only these chapters and MkDocs; the runnable application lives in a separate codebase (see Getting started).

The material is provided as-is: when exchanges, APIs, or libraries change, use when things change — adapting (paths refer to the full Bottie project layout).

Never used Python or unsure what an “API” is? Start with Chapter 0 — New to Python and APIs (where to install Python, what an API is, and where to get Binance / Finnhub keys on the real websites). It is written to be self-contained.

Publishing or selling this material? See Note for creators — pricing and donations.

Hosted docs (MkDocs): https://newbot.berta.one

The folder layout mirrors the Berta Chapters idea: each chapter has a README.md as the spine, room for Jupyter notebooks, exercises, and pointers into the production code in the application repository. You can still use Berta AI to draft additional notebooks or chapters and merge them here the same way you grow the main Berta curriculum.

Disclaimer: This material is software engineering education, not investment, tax, or legal advice. Cryptocurrency markets are volatile. If you ever decide to use it for live trading or real funds, start with testnet and paper mode, keep size small, and only scale once you trust your setup and your own risk controls.

For context on expectations vs how markets actually work (including why an AI assistant is not a substitute for discipline, infrastructure, and risk management), see Luigi’s article: Claude didn’t earn $10k overnight — and that isn’t how markets work.

Who this is for

  • Complete beginners — after Chapter 0, you can continue even if you had never installed Python before.
  • Developers who want a guided tour of a non-trivial Python trading stack (not a toy script).
  • Former users of Bottie who want a structured reference to teach or sell a “build it yourself” product without appearing on camera.
  • Students of your Berta-style pedagogy who expect objectives → reading → labs → exercises → troubleshooting.

If you already code: you should be comfortable with Python, virtual environments, and HTTP basics (GET/POST, JSON). If not, Chapter 0 + Getting started close the gap.

What you will be able to do

After working through the curriculum (and doing the labs, not only skimming):

  1. Explain the path from an HTTP request to optional trade execution.
  2. Run render_api.py locally and interpret /health and related payloads.
  3. Run scripts/smoke.py and map log lines to engine/ modules.
  4. Configure environment variables safely (no secrets in git).
  5. Extend or replace the ML probability layer without breaking the signal pipeline.
  6. Decide when quantum and sentiment features are worth the operational complexity.

How long it takes

Path Chapters Guided hours (read + labs)
API + deploy focus 1, 2, 3, 7 ~12–15 h
Signals + paper trading 1, 2, 4, 5, 7 ~18–22 h
Full linear track 1 → 7 ~30–40 h

The “guided hours” assume you type the commands, break things once or twice, and read error messages. Power users go faster; first-time FastAPI learners should double the estimate for Chapters 2–4.

Start here

Step Link
Install, env, first commands Getting started
Full syllabus and learning paths Curriculum
Big-picture architecture Chapter 1 — Overview
Hands-on environment Chapter 2 — Environment
Goal Where
Official HTTP entry (Render / local) render_api.py
Alternate / legacy monolith API api/app.py
Bar → signal → weights → executor engine/
Batch probabilities for signals ml/ensemble_model.py
Paper integration smoke (no exchange keys required for basic path) scripts/smoke.py
Automated tests pytest tests/

Glossary (short)

Term Meaning in this repo
Bar One candle / tick summary: symbol, OHLCV-style fields consumed by SignalGenerator.
Paper Simulated execution (no real exchange orders), controlled via executor / env.
Quote Usually USDT (or configured quote currency) for pair symbols like BTC/USDT.
Smoke scripts/smoke.py — synthetic bars through the integrated engine to verify imports and control flow.

Building this site locally

See Getting started to build this documentation site locally.

Structure inspired by Berta Chapters; Bottie-specific content is maintained in this repository.