Session-aware, roll-adjusted futures data with database-backed provenance and quality controls.
Quantitative software architecture for disciplined futures research and execution.
Dune Point Technologies develops a private Trade Framework spanning futures data engineering, systematic strategy research, shared signal semantics, portfolio-aware allocation, and broker-connected execution control.
Framework positioning
An integrated institutional-style software package, not a retail trading screen.
The Trade Framework connects the full lifecycle of systematic futures development: raw vendor history becomes auditable research data, candidate systems are discovered inside governed constraints, approved logic is promoted through a canonical signal contract, and live operation is controlled through allocation, reconciliation, durable state, and fail-closed supervision.
Multi-objective discovery, realistic simulation, layered validation, and reportable analytics.
A shared feature-generation contract reduces drift between research evidence and live behavior.
Broker-aware execution, portfolio gates, reconciliation, supervision, and auditable state management.
Software packages
Three specialized systems coordinated by one operating model.
Each package has a clear responsibility boundary. That separation keeps data preparation, strategy research, and live execution governance independently auditable while preserving a coherent path from research artifacts to runtime decisions.
Data Management Software
Transforms third-party futures files into normalized, quality-controlled, session-aware, continuous contracts suitable for systematic research and downstream trading-system workflows.
- Contract-level and continuous-symbol data management
- Roll-event computation and additive adjustment
- MySQL-backed ingestion, audit, and usability reporting
Quantitative Research Software
Converts a governed strategy design space into candidate systems through repeatable discovery, realistic trade simulation, robustness validation, and institutional-style reporting.
- Multi-objective genetic system discovery
- Out-of-sample, walk-forward, and cross-market validation
- Deterministic diagnostics, Monte Carlo analysis, and PDF reporting
Execution Software
Bridges validated research output into broker-aware operation with market-data readiness checks, order lifecycle controls, portfolio admission gates, reconciliation, and supervision.
- Shared signal semantics between research and execution
- Bid/ask-aware trigger evaluation and order management
- Runtime supervision, fault detection, and durable audit evidence
Methodology
The principal strength is lifecycle coherence.
Data quality precedes model quality
The platform treats timestamp policy, contract lineage, roll behavior, missing data, and usability ranges as primary engineering concerns instead of assuming vendor history is directly research-ready.
Discovery is constrained and governed
Candidate systems are searched inside feasible configuration boundaries, evaluated against competing objectives, and filtered through robustness workflows before promotion.
Simulation is execution-aware
Research evaluation accounts for costs, slippage, session behavior, trade timing, position accounting, and reporting artifacts rather than relying only on simplified bar-close assumptions.
Runtime decisions are auditable
Live execution is surrounded by state persistence, broker-truth reconciliation, runtime health checks, and explicit control boundaries.
Architecture narrative
Public architecture without exposing proprietary implementation detail.
The public description remains at the methodology and package-boundary level. It does not publish source code, strategy definitions, database credentials, live positions, broker account identifiers, parameter maps, private endpoints, or trade recommendations.
Important public boundary
Dune Point Technologies is not offering investment advisory services, brokerage services, trading signals, managed accounts, or public access to the private Trade Framework through this website.