SiStarAI / QUANT RESEARCH & TRADING
AIQuantMind | Quant research and trading platform
Move from a research hypothesis to an inspectable strategy workflow. Test rules against history, observe operation in simulation, then confirm account authorization, risk constraints and human approval before live use.
Published by FiClaw 产品团队 · Published · Updated
阅读中文版What is AIQuantMind?
AIQuantMind is SiStarAI's quant research and trading platform, covering strategy development, historical backtesting, AI-assisted strategy iteration, paper trading, live integration and risk controls. FiClaw focuses on natural-language research tasks, strategy code and experiment interpretation; AIQuantMind provides the quant engine and strategy runtime. They work together through quant service interfaces. Modules, data, account permissions and deployment are confirmed for each enabled environment or project.
Core capabilities
01
Strategy development and AI assistance
Organize strategy code, conversion and correction tasks, using backtest feedback to refine rules and parameters. Researchers confirm intent and the scope of changes.
02
Historical backtesting and analysis
Run strategies within supported data coverage and inspect returns, drawdowns and trade records. Retain code, settings and data conventions for review. This introduction primarily addresses A-shares.
03
Parameter comparison and iteration
Compare candidate rules and parameters, retaining changes and experimental evidence. Optimization still requires out-of-sample validation and human judgment.
04
Paper trading
Observe signals, simulated matching, positions and performance in a continuously running environment. Simulation results do not establish actual live execution.
05
Live and account integration
Connect configured trading adapters and authorized accounts for orders, positions and execution queries. Account scope, broker interfaces and permitted actions require explicit confirmation.
06
Risk controls and operating records
Use enabled risk modules for trading constraints, blacklists and review records, retaining runtime and exception information. Institutions define rule scope and response responsibilities.
AIQuantMind use cases
Explore common questions in quant research, backtesting and strategy operation.
A-share strategy backtesting
How can a team evaluate stock selection or rebalancing rules?
Define the universe, signals, period and transaction costs. Compare historical returns, maximum drawdown, Sharpe ratio and trade records, then inspect out-of-sample results.
Review a backtest report →AI-assisted strategy iteration
How can backtest results inform the next experiment?
Describe a hypothesis and organize strategy code with FiClaw, then compare historical rules and parameters in AIQuantMind. Keep changes and evidence for researchers to review.
Explore parameter optimization →From backtesting to paper trading
How can a team observe ongoing strategy operation?
Observe signals, simulated orders and positions, then review settings and exceptions. Before live access, confirm the account, trading interface and risk constraints.
Explore quant capabilities →Private fund and institutional research
How can a quant team review strategies together?
Organize rules, data conventions, parameters and results so researchers can discuss candidates. Team access and deployment are scoped to actual requirements.
Explore private fund research →Review each stage before progressing
01
Define the research task
Specify universe, signals, rebalance, costs, information timing and the hypothesis.
Evidence: Rules, data scope and open questions
02
Backtest and compare
Run candidates, inspect trades and differences, then decide whether to revise.
Evidence: Code, parameters, results and review
03
Observe paper operation
Review continuous operation, signals, positions and exceptions against research expectations.
Evidence: Environment settings, simulated orders and logs
04
Authorize live access
Confirm accounts, interfaces, constraints, launch approval and exception procedures.
Evidence: Authorization, limits, release record and owners
How AIQuantMind works with FiClaw
FiClaw
AI research workstation
Organizes natural-language tasks, specifications, code, backtest diagnosis and human review.
AIQuantMind
Quant research and trading platform
Provides backtesting, paper and live strategy operation within enabled modules, data and account settings.
Quant service interfaces
Connect tasks and underlying capabilities
FiClaw uses services such as FiClawQuantAPI to access quant capabilities. Confirm the actual integration path for the project.
Access and deployment
Scope around the team's task
Discuss research validation, simulation or authorized trading integration. Define modules, data permissions, users and service scope explicitly.
Assess deployment and connections
Evaluate private environments, sources, accounts and trading interfaces per project. Agree licensing, operations, maintenance, pricing and delivery scope.
Retain human decisions and review
Downloading FiClaw Desktop does not automatically enable AIQuantMind live access. Accept historical testing, simulation and live use separately; authorized people retain launch and capital decisions.
Common questions
How are AIQuantMind, AIQuant and AiQuant related?
This page uses AIQuantMind for the platform introduction. Related interfaces and engineering material also use AIQuant and AiQuant for frontend and engine-related modules. These names do not establish three independently purchased products. Confirm versions and modules in the delivery scope.
How does AIQuantMind differ from FiClaw?
FiClaw is an AI workstation for research tasks. AIQuantMind provides the quant platform for backtesting, simulation and trading operation. They can work together through service interfaces; purchasing and enabled scope depend on the agreed solution.
Does downloading FiClaw enable live trading?
No. Live use needs separately confirmed account authorization, interfaces, risk constraints and human approval. Passing research tests or simulation does not grant launch approval.
Which markets, accounts and deployments are supported?
This introduction primarily addresses A-share research and trading. Other markets, account counts, broker interfaces, historical coverage and private deployment require assessment against actual modules and integration conditions.
Can AIQuantMind backtest A-share strategies?
The platform supports historical A-share strategy research within enabled data coverage. Researchers can inspect rules, trade records and results; historical coverage and permissions depend on access.
How does backtesting differ from paper trading?
A backtest evaluates rules on historical data. Paper trading observes ongoing strategy operation and simulated orders. Neither replaces validation with an authorized live account and trading interface.
How should a quant backtest be reviewed?
Examine returns, maximum drawdown, Sharpe ratio and trades alongside the universe, data timing, costs and out-of-sample results. No single metric establishes sustained profitability.
Do optimization or historical tests guarantee returns?
No. Tests evaluate rules under historical data and assumptions. Simulation and live results also depend on data, costs, execution, systems and changing markets. Authorized people remain responsible for decisions.
Explore the platform or discuss your team workflow
Use the platform entry point with an authorized account. For a team evaluation, describe the research task, required modules and integration scope.