Risk governed
Pre-trade, in-trade, and post-trade controls keep execution risk outside strategy code.
Trading OS for Taiwan Index Futures
Taifex Quant Trading Platform
Turn TX / MTX / TMF quant work from scattered scripts into a governed Trading OS.
A Web command center for data governance, strategy research, reproducible backtests, paper and shadow workflows, risk controls, OMS state management, broker-gateway isolation, and future auditability.
Operationally visible
Command Center, audit logs, and observability turn quant workflows from black boxes into managed systems.
Paper-first
Current defaults remain paper-only while the platform foundation, controls, and review process mature.
Browser-only Paper Trading demo: no broker, no real order, no credentials, and not investment advice.
Research
Backtest
Paper
Shadow
Small Live
Full Live
Current defaults stop at paper trading until future approval, governance, and controls are added.
Platform thesis
Trading infrastructure for a focused market
Taifex Quant Trading Platform is not a simple trading bot. It is a trading operating system for Taiwan Index Futures workflows where data governance, strategy research, backtesting, risk control, order management, broker isolation, and auditability need to work as one governed stack.
Individual traders and smaller desks often stitch these capabilities together manually with scripts, spreadsheets, broker tools, and local databases. This project aims to turn that fragmented workflow into a scalable SaaS and enterprise platform foundation.
Business positioning
From MVP scripts to a Trading OS
The commercial plan positions the platform as a durable trading operating system rather than a one-off automation script. The website presents the business plan as decision-ready product architecture: a modular stack, governed execution boundaries, and enterprise continuity across data, risk, OMS, and broker adapters.
Technical foundation
Traditional script stack
Local Python scripts, ad hoc files, single-machine operations
Trading OS platform
Containerized services, shared storage, service boundaries, replaceable adapters
Business impact
Supports multi-user workflows, team operations, and future enterprise deployment paths
Data governance
Traditional script stack
CSV exports and unreconciled local databases
Trading OS platform
Versioned market data, rollover-aware datasets, analytics storage, reproducible inputs
Business impact
Reduces backtest-to-live drift and improves institutional trust
Risk control
Traditional script stack
Hard-coded checks inside a strategy script
Trading OS platform
Dedicated risk engine policies before OMS and broker gateway
Business impact
Creates a compliance-oriented control plane instead of hidden assumptions
Audit and continuity
Traditional script stack
Local logs and operator memory
Trading OS platform
Centralized events, role boundaries, recovery-oriented architecture
Business impact
Improves operational resilience and partner confidence
Target instruments
TX / MTX / TMF exposure normalization
Taiwan futures sizing needs a shared risk language. TX-equivalent exposure keeps strategy limits, portfolio checks, and paper-to-live controls aligned across contract sizes.
TX
NTD 200
Per index point
MTX
NTD 50
Per index point
TMF
NTD 10
Per index point
1 TX = 4 MTX = 20 TMF 1 MTX = 5 TMF
Architecture overview
A platform, not a single trading bot
The system separates command surfaces, research workflows, risk decisions, order management, broker access, and analytics storage so future live execution can be governed and audited.
Web Frontend
API / Backend
Strategy Registry Data Pipeline Risk Engine OMS Broker Gateway
Future Event Bus / Observability
PostgreSQL Redis ClickHouse Data Lake Future
Safety-first trading design
Paper-first by default
The repository is designed to keep execution risk constrained while the platform foundation is built. Live trading requires explicit future approval, additional controls, and compliance review.
TRADING_MODE=paperENABLE_LIVE_TRADING=falseBROKER_PROVIDER=paper Strategy/execution separation
Strategies emit signals only. They do not call broker SDKs directly.
Risk before order execution
Orders must pass through the Risk Engine and OMS before any broker gateway boundary.
No secrets in source
Broker credentials, account IDs, API keys, and certificates must not be committed.
Feature architecture
The operating system layer behind systematic Taiwan futures trading
The platform is designed as an integrated workflow: govern the data, research the strategy, validate in paper or shadow mode, then keep every future order behind Risk Engine, OMS, and Broker Gateway boundaries.
Data foundation
Market Data Pipeline
Ingest, validate, version, and serve TX / MTX / TMF bars, ticks, contract master data, and quality reports.
Reduces data errors that distort research and backtests.
Data foundation
Rollover Engine
Separate research-only adjusted continuous futures from real-contract prices used for paper or future execution simulation.
Addresses the classic backtest/live drift caused by contract rollover.
Research
Strategy Lab
Bind strategy versions, data versions, research contexts, and backtest artifacts into reproducible review packets.
Makes strategy results reviewable instead of anecdotal.
Validation
Paper and Shadow Trading
Validate behavior in paper-first and future shadow workflows before any broker-bound path is considered.
Creates a safer path from research to operating readiness.
Control
Risk Engine
Centralize exposure limits, stale quote checks, daily loss controls, duplicate prevention, and future kill-switch policy.
Keeps high-risk decisions in a dedicated control layer.
Control
OMS
Own deterministic order state, idempotency keys, event-style transitions, and future reconciliation inputs.
Turns order handling into an auditable process.
Execution boundary
Broker Gateway
Isolate broker adapters behind a normalized gateway so strategies never call broker SDKs directly.
Reduces vendor coupling and protects the strategy layer.
Operations
Web Command Center
Provide a command surface for mode visibility, safety flags, review packets, health, and future manual controls.
Gives operators one place to inspect system state before decisions.
Enterprise
Audit and Observability
Plan for OpenTelemetry traces, immutable audit records, recovery views, and incident review workflows.
Supports institutional review, business continuity, and compliance conversations.
Future analysis
AI-Assisted Diagnostics
Future-facing support for overfitting checks, experiment summaries, anomaly review, and strategy documentation.
Speeds research review without turning AI into a trading authority.
Commercial design
Recurring software revenue before regulated execution services
The business plan translates into a staged revenue model that starts with software subscriptions, usage-based analytics, and enterprise delivery instead of premature claims about trading profitability.
Illustrative pricing and packaging from the business plan are presented as directional commercial design, not a public offer.
Monetization paths with service logic
SaaS subscriptions
Service
Self-serve research, backtesting, paper trading, dashboards, alerts, and workflow storage.
Business logic
Recurring revenue is tied to workflow depth, data precision, compute capacity, and retained operating history rather than trading outcomes.
Use cases
- Beginner quant onboarding
- Individual research workspace
- Paper-trading workflow validation
Value
Creates predictable ARR while giving users a safe path from research to paper operation.
Research tooling and paper trading only by default.
Professional trader plans
Service
Higher-frequency data access, webhook risk alerts, strategy runner capacity, richer reporting, and production-like paper workflows.
Business logic
Pricing expands with operational maturity: second-level or tick workflows, more compute, higher retention, and better monitoring.
Use cases
- Active discretionary-plus-systematic traders
- Small teams validating strategies
- Shadow trading preparation
Value
Helps serious users reduce manual workflow gaps without promising investment performance.
No live order routing by default; broker access remains a reviewed future boundary.
Enterprise licensing
Service
Private cloud or on-prem deployment, RBAC/ABAC direction, WORM audit log direction, SLA, security review support, and custom integration scope.
Business logic
Enterprise revenue is based on governance, deployment control, auditability, support terms, and integration complexity.
Use cases
- Proprietary desks
- Family offices
- Broker innovation teams
- Institutional research groups
Value
Turns the platform into procurement-ready infrastructure with clearer risk ownership and operating boundaries.
Contracts should define platform responsibility, customer trading decisions, audit retention, and incident handling.
Data services
Service
Rollover-aware TX/MTX/TMF datasets, cleaned bars and ticks, data quality reports, research-only adjusted continuous futures, and data APIs.
Business logic
Cleaned and versioned Taiwan futures data can become a high-retention asset because research, backtests, and operations depend on consistent inputs.
Use cases
- Backtest data packs
- Data API subscriptions
- Enterprise data governance
- Rollover validation reports
Value
Improves reproducibility and reduces backtest-to-live drift for users who currently rely on scattered files.
Data licensing, redistribution rights, and exchange/vendor restrictions must be reviewed.
Strategy marketplace
Service
Future marketplace infrastructure for strategy authors, validation reports, risk labels, versioned strategy packages, and review workflows.
Business logic
Marketplace economics can create network effects once the platform has review, risk labeling, and clear author responsibility.
Use cases
- Strategy author distribution
- Research template library
- Institutional strategy review queue
Value
Connects developers and users around standardized signal contracts without letting strategies bypass risk or OMS.
Copy trading, signal subscriptions, or advisory-like distribution require separate legal and regulatory review.
AI analysis add-ons
Service
AI-assisted overfitting checks, parameter sensitivity summaries, anomaly review, strategy documentation, and operational diagnostics.
Business logic
Usage-based AI diagnostics can monetize compute-heavy review workflows while keeping human users in control.
Use cases
- Backtest review
- Strategy health reports
- Experiment comparison
- Incident postmortem summaries
Value
Speeds research review and operational learning without turning AI into an autonomous trading authority.
AI output must remain analytical tooling, not individualized investment advice.
Broker or institutional partnerships
Service
White-label infrastructure, co-branded research tools, broker adapter projects, training programs, and institutional workflow pilots.
Business logic
Partner revenue comes from integration, distribution, and operational enablement rather than unreviewed order-routing economics.
Use cases
- Broker developer portal
- Fintech sandbox
- Institutional pilot deployment
- Education and onboarding programs
Value
Gives partners a governed Taiwan futures quant stack while preserving broker-gateway isolation.
Broker fee-sharing, referrals, or order-routing monetization are compliance-dependent future options.
Compliance-dependent performance models
Service
Future-only structures such as performance fees, managed accounts, copy trading, signal subscriptions, or broker fee-sharing.
Business logic
These can only be considered after licensing, legal review, customer suitability, operational controls, disclosure, and conflict management are defined.
Use cases
- Licensed partner structure
- Reviewed signal distribution
- Approved managed-account framework
Value
Preserves optionality without marketing regulated services as currently available.
Not available in the current product; no profit guarantee and no live trading by default.
Illustrative subscription tiers
Basic
Quant beginners
~TWD 1,980 / month
1m data
- Core backtesting
- Paper trading
- Foundational workflow visibility
Pro
Active traders
~TWD 8,800 / month
1s data
- API access
- Webhook risk alerts
- Higher-frequency workflow support
Elite
Professional traders
~TWD 29,800 / month
Tick data
- Multi-account reconciliation
- AI strategy diagnostics
- Cloud runner capacity
Enterprise
Institutions and family offices
TWD 250,000+ / month
Level 2 depth
- Private cloud deployment
- Audit-oriented logging
- SLA-backed integration
Usage-based expansion
AI diagnosis tokens
High-margin analysis credits for overfitting detection, parameter sensitivity review, and strategy health checks.
Tick replay and TCA workloads
Additional billing for high-frequency replays, historical reconstruction, and execution-cost analysis workflows.
Enterprise delivery
Private cloud and on-prem delivery
Longer-term enterprise licensing with deployment control, data locality, and institution-specific operating boundaries.
Adapter and workflow integration
Paid integration work for broker adapters, internal controls, reporting, and operational change management.
Performance fees, managed accounts, copy trading, signal subscriptions, or broker fee-sharing may require legal, regulatory, or licensed partner review.
Customer strategy
Pricing and messaging by operating maturity
The business plan distinguishes between self-directed traders seeking better local-market tooling and institutions buying governance, continuity, and control. The website reflects that split directly.
Retail and professional traders
Lead with localized data quality, rollover-aware datasets, and a clear research-to-paper workflow.
- TX / MTX / TMF risk-equivalent sizing
- Continuous futures data as a premium differentiator
- Trading journal and workflow stickiness for retention
Institutions and family offices
Sell governance capabilities, role separation, and infrastructure resilience rather than a promise of strategy alpha.
- RBAC and operating boundary design
- High-availability deployment direction
- Auditability, reconciliation, and procurement readiness
Go-to-market and moat
Specialize first, expand from infrastructure
The business plan argues for a Taiwan futures-first wedge. The moat is not a single strategy; it is the accumulated operating system around localized data, controlled execution, repeatable workflows, and ecosystem gravity.
Three-stage growth path
Phase 1: Tool-led growth
Acquire early technical users through a strong backtesting foundation and better rollover-aware market data.
Phase 2: Ecosystem expansion
Add broker adapters, collaboration workflows, and a strategy marketplace to lower acquisition cost through third-party participation.
Phase 3: Infrastructure moat
Move upmarket with enterprise controls, deployment flexibility, and stronger audit and governance expectations.
Durable advantages
Proprietary Taiwan futures datasets
Versioned, cleaned, rollover-aware datasets become operational infrastructure rather than commodity files.
Broker abstraction layer
A neutral broker gateway model reduces lock-in to any single broker and supports continuity planning.
Workflow and audit history
Backtest, paper, risk, and order-state history compound into a durable switching cost for teams.
Roadmap
Realistic platform buildout
The current repository is a development skeleton. The roadmap progresses from foundation to governed execution boundaries before any future live workflow.
- Phase 1: Development foundation
- Phase 2: Strategy signal contract and TX/MTX/TMF risk sizing
- Phase 3: Data pipeline and rollover engine
- Phase 4: Backtesting and paper trading
- Phase 5: Risk Engine and OMS skeleton
- Phase 6: Broker gateway integration
- Phase 7: Web command center
- Phase 8: Enterprise controls, audit, observability, deployment
Governance and compliance
Compliance framing as a commercial capability
The business plan treats compliance as a sales accelerator. Clear system boundaries, security alignment, and operational responsibility definitions reduce friction with institutional customers and partners.
SaaS-first commercial boundary
The early platform position is software and infrastructure tooling. Regulated trading services remain outside the default product scope.
Procurement readiness
Security controls, audit trails, role separation, and kill-switch design shorten third-party risk review for larger customers.
Contract clarity
Reconciliation, service levels, and emergency controls should be explicit in customer agreements, not implied by marketing copy.
Compliance and risk notice
Research software, not investment advice
This website and repository are for research, engineering, and product development. They do not provide investment advice, do not guarantee profit, and do not remove the substantial risk involved in futures trading.
Live trading, signal services, managed accounts, copy trading, broker fee-sharing, or other regulated services require separate legal and compliance review before any release or customer use.
Research and engineering development
Build the platform foundation before execution risk
Open the browser-only Web App demo, review the safety principles, and continue in small verified slices.