EDGE-OPT-04 · Complete · private production
Options Edge
A completed, simulation-only command center that turns live market evidence into structured, explainable options decisions while keeping execution human-controlled.The challenge
Options workflows produced abundant market data without a disciplined, explainable path to a decision.
Options analysis creates an abundance of data but not necessarily clarity. Options Edge converts live market evidence into a disciplined review workflow while preserving manual execution and explicit human judgment.
The approach
Break the decision into evidence, rules, and authority.
- 01
Converted the trading thesis into explicit setup, contract, quality, and economics gates.
- 02
Placed contextual AI after deterministic market evaluation.
- 03
Kept brokerage execution manual and outside the application boundary.
Architecture / system
A connected operating system for the decision.
- Live read-only Schwab market data and five-minute price history
- Options-chain screening with deterministic setup, contract, quality, and execution-economics gates
- Context-aware AI conversation, manual trade intake, and active-trade monitoring
- Long Shot research plus replay, calibration, and audit tools
Product experience
AI design
What the model does
Supports contextual questions, research synthesis, and explanation of review states.
What remains explicit
Market screening, policy gates, sizing constraints, and simulated accounting remain explicit and reproducible.
Where judgment remains
The user reviews the evidence and performs any trade manually outside the system.
Outcome
Converted noisy options data into a gated review workflow that explains why a setup advances, fails, or needs human judgment—without automating execution.
- Live read-only market data
- Deterministic quality gates
- Manual execution boundary
Project facts:Stack: Schwab API and OAuth, OpenAI API, deterministic policy gates, encrypted token storage, and a simulation-only execution model. Role: product strategy, decision-system architecture, AI experience design, and safety-boundary design. Outcome: delivered an end-to-end decision-support application that produces explainable review states while keeping execution manual. A credential-free public demo at /demos/options-edge uses deterministic simulated data, not live Schwab data; production remains private.
07 · Lesson / perspective
Decision support is strongest when it can explain why an opportunity failed, not merely rank what survived.