QuantComputational

Autonomous SPY Options Trading App
with deterministic filtering + AI strategy selection

This is a Python + Streamlit trading platform that ingests live SPY option chain data (SchwabDev API), applies deterministic policy gates and ranking, and uses an LLM (OpenAI) to select and justify a strategy. The design emphasizes auditability, modular architecture, and traceable decision artifacts.

View GitHub Repo Live Project Stats Demo

Highlights

Deterministic option filtering (DTE, delta band, liquidity, spread ratio) with auditable criteria snapshots
Payload size reduction and schema normalization before LLM calls
Separation of concerns: enrichment vs gating vs ranking vs LLM strategy selection
Structured logging and raw prompt/response captures for traceability

Live Project Stats

Auto-updated from a private build pipeline. No source code is published here.

Loading activity…

Total Code Lines (LOC)

—

Commits (total)

—

Commits (last 30 days)

—

Last Commit

—

Repo Snapshot

commit: —

What I built

Option chain ingestion and preprocessing with deterministic scoring
LLM strategy selection prompt system with raw prompt/response logging
Risk-aware gating policies (spread, liquidity, DTE, delta bands) designed for auditability
Streamlit UI for monitoring, inspection, and iterative refinement
Test scaffolding for key modules to prevent regression and drift

Tech stack

Python, Streamlit, OpenAI API, SchwabDev API, pytest, structured JSON logging.

Demo

Add a short video walkthrough (recommended: YouTube unlisted or Vimeo). Embedding keeps hosting fast and low-cost.

Tip: keep video under 3–5 minutes. Recruiters actually watch short demos.

Language Breakdown

Pulled from stats.json generated by automated tooling.

Language	Files	Code Lines	Comment Lines	Blank Lines
Loading…

Autonomous SPY Options Trading Appwith deterministic filtering + AI strategy selection