Structure inputs before rules
The design uses explicit Product DNA and classification fields so the planning context can be reviewed before any applicability path is evaluated.
Project
Compliance Plan Generator
An in-development, rule-driven system for structuring product inputs, applicability logic, and traceable compliance-planning outputs with mandatory human review.
Project thesis
This in-development system is intended to turn structured product and market inputs into traceable draft compliance-planning outputs. Applicability remains deterministic, unresolved questions stay visible, and expert review is mandatory.
Project scale
- Development focus
- Structure, traceability, and review
- Decision boundary
- Deterministic applicability; AI does not decide scope
- Public boundary
- Architecture only; no public demonstration
Context and constraints
Compliance planning requires product characteristics, target markets, regulatory domains, missing information, applicability rules, required evidence, and review decisions to remain coordinated. Those inputs can be incomplete and rules can change by market, while proprietary knowledge must stay protected and every conclusion still requires expert judgment.
Selected engineering decisions
The current architecture is designed to move from structured Product DNA and classification through deterministic applicability rules, requirement packages, and draft outputs. AI may support explanation or drafting, but it does not determine final applicability. Missing inputs and open questions remain explicit, with human review serving as a required system boundary.
Separate decisions from drafting
Applicability remains deterministic and traceable; AI assistance may explain or draft text but cannot establish the final compliance scope.
Make expert review a system boundary
Carry assumptions, missing inputs, and open questions into review instead of hiding them behind an apparently final plan.
System model
Structured product inputs
The current design starts with Product DNA and target-market inputs that establish the planning context.
Product classification
Classification is intended to organize the attributes that may affect later rule evaluation.
Applicability rules
Deterministic rule paths are being developed without exposing proprietary rule content.
Requirement packages
The architecture is intended to group related requirements without implying complete rule coverage.
Structured draft output
Draft output structures are intended to carry source inputs, assumptions, and planning context forward.
Missing inputs and open questions
Unresolved information remains visible for follow-up instead of being converted into an unsupported conclusion.
Human review
Expert review remains mandatory before any draft output can inform an operational decision.
Interface evidence
Outcome and current boundaries
The project remains in active development, with no public demonstration and no readiness for external operational use. Rule coverage, output structures, and review behavior are still being implemented and validated. This public record describes architecture and workflow only; proprietary rules, internal datasets, and company-sensitive material are excluded. Any output remains draft material requiring expert review.