← Projects

Quant engine architecture

Q-Bet

A modular Python quant engine for matched betting, sports arbitrage, strategy simulation and controlled execution planning.

Independent project - core domain stage
  • Python
  • Pydantic
  • pytest
  • Quant modeling
  • Data pipelines
Target product

A cloud-ready quant betting portfolio

The intended v1 is a web application with a complete matched-betting engine, simulation mode, capital orchestration, collectors or API adapters, reports and controlled execution. The useful angle is not a betting promise, but a system that turns manual opportunity work into modeled, testable workflows.

Current state

Domain models first, execution later

The repo is currently in the core domain stage: typed models for events, markets, offers, opportunities, strategy results and execution plans are implemented and ready for review. Matched-betting calculations, orchestrator logic, collectors, UI and execution adapters come after the domain layer is stable.

Portfolio signal

Math, risk and software boundaries

Q-Bet should communicate modeling discipline: expected value, ROI, liquidity, risk, capital lock-up and approval boundaries need to become explicit software objects before any real execution is even considered.

Engine progression

From odds data to controlled execution

The page should show the future system without overselling the current implementation. The active work is the left side of the flow; the v1 target is the full chain.

01Events / markets / offers
02Typed domain layer
03Opportunity detection
04Strategy calculation
05Risk & capital rules
06Simulation mode
07User-approved execution
08Reports / dashboard

Roadmap

Current stage to v1 target

Current

Stage 1

Core Pydantic domain models and tests for the objects the engines will rely on.

Next

Base Engine

Matched-betting calculations, free-bet strategies, dutching, stake optimization and arbitrage checks.

Planned

Capital Layer

An orchestrator that routes capital by EV, ROI, risk, liquidity and lock-up across strategy engines.

Target

Cloud v1

Web UI, collectors, API adapters, simulation controls, performance reports and approved execution flows.

Portfolio angle

What this page should prove

  • Q-Bet is early, so the page must clearly separate implemented domain layer from future v1 target.
  • The engine flow should emphasize calculations, constraints and approval boundaries, not hype around automation.
  • Later portfolio assets should show formulas, example opportunity objects and simulation output.