ASTRIVERA Platform

Automated Trading Operations and Management Platform

Product overview

ASTRIVERA Platform is an independently developed automated trading infrastructure built on a component-based strategy architecture. It brings strategy logic, portfolio management, risk controls and trade execution into a coherent operating framework.

Strategies, models and functional components interact through standardised interfaces and can be configured, combined and versioned independently. Component management and strategy assembly allow trading capabilities to expand within well-defined architectural boundaries, with controlled verification supporting each release. This approach balances flexibility, operational stability and long-term maintainability.

The platform builds trading plans from strategy objectives, account holdings, available capital and applicable market rules. Capital, position, trading authority and risk constraints are enforced before execution. During trading, it manages order submission, status transitions, execution reports, positions and account information, while identifying, blocking and reconciling exceptional states within a controlled execution process.

An independent quantitative immunity mechanism monitors strategy behaviour, risk characteristics and system health. It identifies deteriorating strategy effectiveness, changes in market suitability and potential signs of strategy failure. It also detects discrepancies across data, signals, execution and accounting, providing evidence for strategy review, risk containment and incident response.

Compliance controls and auditability are embedded throughout trading operations. The platform incorporates broker access permissions, execution authorisation, pre-trade risk checks and operational traceability in support of institutional programme-trading requirements. Structured event logs and audit records link data versions, strategy decisions, risk outcomes, submitted orders, execution reports and account reconciliation, giving institutions a traceable basis for trading reviews, incident investigations and compliance oversight.

Security and resilience begin with clearly defined system boundaries. Local deployment, network isolation by default and least-privilege design restrict unnecessary external access. Authorised communication paths remain available for market-data acquisition and broker connectivity. Strategy computation, risk controls, operating records and diagnostics run independently within a controlled local environment, reducing dependencies on external services and safeguarding critical operations and data assets.

Runtime monitoring, execution lifecycle management, order and account reconciliation and recovery procedures support operational continuity and consistency. Critical execution actions have identifiable states and ownership, enabling independent incident review, controlled recovery and historical traceability.

Core Capabilities

Component-Based Strategy Architecture and Versioning
Multi-Strategy and Portfolio Management
Automated Trade Execution
Portfolio and Risk Controls
Broker Connectivity and Execution Authorisation
Quantitative Immunity and Strategy Degradation Monitoring
Trading Compliance and Access Management
End-to-End Logs and Audit Trails
Order, Fill and Account Reconciliation
Local Deployment and Network Isolation
Monitoring, Exception Protection and Recovery

Explore quantitative immunity

Modular assembly

Strategy components, data views, market rules, risk policies, execution modes, broker connections and account binding are managed separately. Each assembly retains its versions and configuration; interfaces, data and account conditions are checked before a run.

Local control plane

Strategy componentsData viewsMarket rules

Risk policiesExecution modes

Broker connectionsAccount binding

Evidence and reportsQuantitative immunity

Strategy sources and delivery

Client strategies

Integrate existing strategies and research models, compare behaviour before and after migration, and configure the required data, risk controls and execution conditions.

ASTRIVERA proprietary strategies

We research strategies for U.S. equities and China A-shares. Each strategy is assessed separately for its market, frequency, risk characteristics, data requirements and deployment conditions.

Commissioned strategy research

Data preparation, strategy development and evaluation follow agreed markets, asset coverage, holding periods and risk constraints. A project may deliver research findings alone, or proceed to system integration and deployment once the agreed validation conditions are met.

Research objectives, deliverables, schedule and fees are agreed separately. Research evaluation and production deployment have separate acceptance processes.

Validated Research outputs

Strategy research outputs validated in Research can be handed to Platform after checking strategy versions, data dependencies and operating conditions. Research validation does not grant execution authorisation.

Data sources

Depending on deployment needs, Platform can use direct external data connections, resources published by DataLake, or a hybrid arrangement. Access permissions, data versions and operating dependencies are checked separately.

Research and runtime

Each run is linked to its strategy version, data, parameters and configuration, making results easier to review and differences easier to compare.

Run mode	Purpose
Backtest	Evaluate historical performance using agreed data, market rules and cost assumptions.
Replay	Reconstruct decisions from frozen inputs to compare behaviour and results.
Dry run	Check assembly and runtime workflows without submitting trading orders to a broker.

Execution and accounts

Generate order plans from portfolio targets, positions, cash and market rules; record broker responses and fills; and reconcile account state after the run.

Brokers and accounts. Connection configurations and individual accounts are managed separately, with explicit permissions, currencies and applicable markets.

Orders and fills. Track submission, acceptance, rejection, partial fills, cancellation and completion. Unconfirmed outcomes remain unresolved.

Positions and cash. Check internal accounting against broker records, retaining discrepancies and the evidence used for reconciliation.

Twin and evidence

The strategy twin checks runtime differences against relevant baselines. Evidence views show the corresponding records and coverage.

Comparison layer	Focus
Decision deviations	Data inputs, signals, target weights and risk decisions.
Execution deviations	Order plans, actual fills, latency, slippage and costs.
Accounting deviations	Positions, cash, cost attribution and account records.
Economic outcomes	Returns, volatility, drawdown, turnover and related measures.

Evidence and reports are linked to specific runs. Uncovered areas are listed separately and excluded from verified conclusions.

Deployment and support

A project may include environment assessment, software installation, data and broker adaptation, strategy integration, validation and operating guidance. Subsequent upgrades, maintenance and support are defined at delivery.

Strategy code, data, credentials and runtime records remain in the client environment by default. New markets, components or account configurations are added subject to the relevant compatibility and authorisation requirements.

Features, support and deployment depend on the current release and the project scope agreed in writing by both parties.

Cross-platform deployment. Supports macOS, Windows and Linux, with local and private deployment configured for institutional environments.

Book a product demonstration

Back to product index