Quant Developer: Architecting Autonomous Trading Systems

Ideas Are Cheap — Execution Is Everything

Applied quantitative finance with a code-first philosophy — a 365-day program building "AutoQuant-Alpha," a complete autonomous trading system integrated with the Alpaca Markets API.

365

day program

core topics covered

Subscribe on Substack →

Free · No card needed · Unsubscribe anytime

From the archive

Oct 1, 2026

Day 108: Bollinger Bands — Calculating Rolling Volatility Bands

The “full recompute” trap

Sep 29, 2026

Day 107 — RSI Math: Coding the Relative Strength Index

The “pandas.rolling().mean()” Trap

Sep 27, 2026

Day 106 — EMA Smoothing: Implementing Wilder’s EMA

The “Standard EMA” Trap

Sep 25, 2026

Day 105 — Multi-Indexing: Managing Multi-Asset Panel Data

The “wide-format dict” trap

Sep 23, 2026

Day 104: Data Merging -- Joining Price and Economic (FRED) Data

The “just merge_asof it on date” trap

Sep 21, 2026

Day 103 — GroupBy Analysis: Aggregating Returns by Market Sector

The groupby().apply() trap

Sep 19, 2026

Day 102 — Rolling Windows: Coding the SMA and EMA Buffers

The df['close'].rolling(20).mean() Trap

Sep 17, 2026

Day 101 — OHLCV Resampling: Converting Ticks to 15m/1h Bars

Week 15, AutoQuant-Alpha: Pandas Time-Series Wrangling

What you'll build

AutoQuant-Alpha

a complete autonomous trading system integrated with the Alpaca Markets API, built across 4 production phases over 365 days

The curriculum

6 core areas covered

01 Algorithmic trading and backtesting strategies

02 Risk management: VaR, CVaR, stress-testing

03 Financial data engineering and API integration

04 Portfolio optimization: Markowitz, Black-Litterman, HRP

05 Real-time WebSocket execution and concurrent systems

06 ML-based alpha factor discovery

What you'll walk away with

AutoQuant-Alpha — a live autonomous trading system, not a backtest notebook

Risk management you've implemented yourself: VaR, CVaR, stress-testing

Daily code commits — momentum and a track record you can point to

What you're signing up for

What makes it different

Every lesson runs against a live trading sandbox, simulating real bid-ask spreads, liquidity constraints, and API rate limits — not a backtest fairy tale.

Format

365 days, structured as 4 production phases, with daily mandatory code commits.

Who it's for

Engineers who want to build and run a real autonomous trading system against live market data — not backtest theory in a notebook.

A few things worth knowing

How often will I get an email?

Frequency isn't fixed — "From the archive" above shows recent issues so you can see the actual pace before you subscribe.

Is this actually free?

Yes — 100% free, no card required, no trial that converts to a charge later.

Can I unsubscribe anytime?

Yes, one click, no hoops — that's a standard Substack feature, not something we gate.

AutoQuant-Alpha doesn't build itself. Day 1 is free.

Join Quant Developer: Architecting Autonomous Trading Systems — free, straight to your inbox. 6+ topics, one subscription.