Options GPS: High-Level System Design

[e35ventura]

Options GPS: High-Level System Design

Category: Best Options Tool
Difficulty: Advanced
Synth Endpoints Used: get_prediction_percentiles(), get_option_pricing(), get_volatility()

---

Product Objective

Turn a trader's view into one clear options decision.

• Inputs are user-driven: bullish / bearish / neutral and low / medium / high risk
• System auto-uses Synth 1h + 24h forecasts — no timeframe question needed from the user

User Input → Output

1. User picks symbol (e.g. BTC, SPY, NVDA)
2. User picks market view (bullish, bearish, neutral)
3. User picks risk tolerance (low, medium, high)
4. App returns 3 strategy cards:

Card	Purpose
Best Match	Highest-scoring strategy for the user's view
Safer Alternative	Higher win probability, lower risk
Higher Upside	Lower probability but larger potential payoff

Each card shows:

• Why it fits the user's view
• Chance of profit
• Max loss
• Invalidation trigger
• "Review again at" time

Core Decision Pipeline

1. Forecast Fusion Engine

• Uses Synth 1h for timing and 24h for session conviction
• Classifies current state as Aligned, Countermove, or Unclear

2. Strategy Generator

• Builds a candidate set based on user bias and risk profile
• Directional: long calls/puts, debit spreads, credit spreads
• Neutral: iron condor, butterfly, defined-risk premium structures

3. Payoff + Probability Engine

• Evaluates each strategy against the Synth probability distribution
• Computes: expected value, probability of profit, tail risk, loss profile

4. Ranking Engine

Scores candidates using a weighted objective:

score = fit_to_view + pop + expected_return - tail_penalty - liquidity_penalty

Weighting shifts based on risk tolerance.

5. Guardrails

• Hard filters for max acceptable loss, minimum liquidity, spread quality, and confidence
• Includes explicit No Trade state when confidence is low or signals conflict

Risk-Tolerance Behavior

Level	Behavior
Low	Prioritize higher win probability, defined-risk structures, tighter loss control
Medium	Balanced probability and payoff
High	Allow lower win-rate, higher convex payoff structures

UX (4 Screens)

Screen 1: View Setup

• Simple two-step input: bias + risk
• Instant summary sentence: "You are moderately bullish with medium risk."

Screen 2: Top Plays

• Three ranked cards with one-line rationale each
• One primary recommendation highlighted

Screen 3: Why This Works

• Distribution view + payoff visualization
• Plain-English explanation of required market behavior

Screen 4: If Wrong

Predefined reroute actions:

• Exit rule
• Convert/roll rule
• Time-based reassessment rule

Technical Architecture

Layer	Responsibility
Data	Synth forecasts + option pricing insights via synth_client
Signal	Fusion engine, candidate builder, scoring/risk engine
Product	Recommendation UI, explainability, alerting
Audit	Decision log — "why this was recommended now"

---

Submission Requirements

• Demo video — Screen recording showing the tool in action
• 1-page technical explanation — Include as tools/options-gps/README.md. Should cover what the tool does, how it works, and how it uses Synth data
• Tests in tools/options-gps/tests/

Getting Started

This repo includes everything you need to build without a Synth API key:

from synth_client import SynthClient

client = SynthClient()  # auto-loads mock data when no API key is set

# Get the probability distribution
forecast = client.get_prediction_percentiles("BTC", horizon="24h")

# Get theoretical option prices
options = client.get_option_pricing("BTC")

# Get volatility metrics
vol = client.get_volatility("BTC", horizon="24h")

1. Fork the repo and clone it
2. Copy the template: cp -r tools/_template tools/options-gps
3. pip install -r requirements.txt
4. Build your tool using the SynthClient — it works immediately with mock data
5. Add tests in tools/options-gps/tests/
6. Open a PR when ready

See the README for the full endpoint reference and supported assets.