From manual trading to tools you understand
Build the tools behind your trading strategy.
Learn to turn your trading rules into software with Python and AI. Start from the beginning, follow one clear step at a time, and use the materials as you build.
The series · 23 episodes
A clear path from idea to application.
Six parts, in order. Videos are coming soon; the first worksheets and setup prompt are ready to use.
Part 1 · Set up and define your strategy
Install Python and set up your AI workspace
Install Python, choose a coding assistant, and check that it can create and run a file in your project folder.
Write your strategy as exact rules
Describe entries, exits, position size and timing so another person could follow the same rules.
Have AI turn one rule into Python, and check it
Turn one plain-English rule into code and check it with examples whose answers you already know.
Part 2 · Build the backtester
Settings and practice data
Create checked settings and load a small practice price dataset for the backtester.
Your strategy becomes signals
Turn your written rules into buy and sell signals you can inspect.
Signal to fill, without peeking at the future
Decide when a signal can become a simulated trade, using only information available at that time.
Positions, costs and tests
Track cash, position size and trading costs, then test that the trade records and totals agree.
Run it on real historical data, and read the evidence
Apply the research workflow to historical prices and look beyond the headline return.
Compare strategies and parameters fairly
Compare settings and strategies without rewarding a lucky fit to the past.
Part 3 · Make it an application
Separate the strategy from the backtester
Give your strategy its own runnable program with clear inputs and outputs.
Give it its own settings file
Make separate strategy applications predictable and independently configurable.
Run it on a schedule during market hours
Start and stop work at defined times, with market sessions and time zones made explicit.
Part 4 · Connect live data
Connect to a live data provider
Connect a strategy application to incoming market prices.
Handle stale prices and duplicates
Recognize old or repeated updates before they can affect a decision.
Recover after a disconnect
Reconnect to a data feed and resume without double-counting events.
Part 5 · Paper trading
Connect and authenticate with a paper broker
Connect a simulated trading account and confirm that the application is using paper trading.
Place orders
Follow a strategy signal through an order request and the broker’s acknowledgement.
Track fills and partial fills
Update the application from what actually filled, including orders filled in pieces.
Keep your positions matching the broker’s
Compare local positions with the paper broker and resolve differences.
Part 6 · Run it safely
Risk limits and a kill switch
Set exposure limits and define how the application stops safely.
Monitoring and event replay
Record what happened and why, then replay events to investigate a problem.
Paper results versus backtest results
Compare paper-trading observations with the assumptions and results of the backtest.
Package it and extend it
Package a working application and add new capabilities without losing a reproducible starting point.
The backtester
Get hands-on while you learn.
Explore the original Applied Quants backtester and see how rules turn into simulated trades. Download the complete practice source to inspect the engine and animated interface.
Free practice version · artificial pricesWork with me
Have a strategy.
Need an application?
I can help turn defined requirements into Python research tools, backtesting workflows, and configurable applications.
- Strategy implementation and customization
- Data workflows and backtest reporting
- Application structure, testing, and troubleshooting
Working together through Upwork
New to Upwork? Discuss your project first; eligible clients can receive a Direct Contract invitation. Already on Upwork? Use my profile to discuss the project.
An ordinary profile link does not activate Direct Contract pricing. Eligibility and fees follow Upwork’s current terms.
Questions about a lesson? Show me what you built. If you find a better prompt or a code improvement, send it over. I’ll review it, test it, and publish an updated bundle if it improves the materials.