Principles & Guidelines

Principles and Guidelines that shape LEF architecture, infrastructure, and platform choices.

---

1. Architecture principles

   BUSINESS RULES ARE FIRST-CLASS ASSETS.

   Sustainable modernization requires business logic to be explicit, governed, and portable. Rules are documented, versioned, testable, and traceable; never only implicit in UI or custom code.

   SECURITY BY DESIGN.

   Trust and continuity depend on security being embedded. Least privilege, secure defaults, secrets management, and auditability are mandatory.

   SEPARATE FUNCTIONAL INTENT FROM TECHNICAL IMPLEMENTATION.

   Technology changes; business intent must remain stable to avoid functional legacy. Model domain/process/rules independently from platforms; technology swaps must not force business redesign.

   EVOLVABILITY IS A PRIMARY QUALITY ATTRIBUTE.

   Long-term success depends on cost of change, not just first delivery. Evaluate architecture on changeability; define migration/refactor paths as part of design.

   MODEL-DRIVEN BY DEFAULT; CUSTOM CODE BY EXCEPTION.

   Modeling increases reuse, speed, and consistency. Prefer Thinkwise modeling/configuration; custom code needs justification, tests, ownership, and an exit path.

2. Business principles

   OUTCOMES OVER OUTPUTS.

   Features are not the goal; measurable impact is. Define success metrics and prioritize what moves them.

   MAKE TRADE-OFFS EXPLICIT.

   Every shortcut has a cost; hidden debt becomes operational pain. Document deviations with rationale, risks, an owner, and an exit path.

   QUALITY IS A STRATEGIC CHOICE.

   High-value systems must be stable, maintainable, and predictable. Quality gates are non-negotiable; “fast” never means “fragile.”

   STANDARDIZE WHAT DOES NOT DIFFERENTIATE.

   Variation increases cost and reduces reliability. Use shared patterns and processes unless a deviation has proven business value.

   DELIVER IN SMALL, HIGH-VALUE INCREMENTS.

   Short cycles reduce risk and increase adoption. Slice work into deployable increments with feedback loops.

3. Information principles

   SINGLE SOURCE OF TRUTH PER DOMAIN.

   Conflicting sources cause operational errors and distrust. One authoritative system per domain; replication is explicit and governed.

   PRIVACY AND COMPLIANCE BY DEFAULT.

   Regulatory and contractual obligations protect clients and LEF. Minimize collection, enforce retention, access control, and auditing.

   DATA IS A PRODUCT WITH AN OWNER.

   Data quality and meaning do not happen automatically. Key datasets have an accountable owner, clear definitions, and quality standards.

   CONTRACT-BASED DATA EXCHANGE.

   Integrations break when changes are unmanaged. Use versioned schemas/contracts, compatibility rules, and deprecation paths.

   TRACEABILITY AND LINEAGE ARE REQUIRED.

   Decisions require trust in numbers. Make origins and transformations of key KPIs traceable and auditable.

4. Technology principles

   SECURITY: LEAST PRIVILEGE AND ZERO TRUST.

   Most incidents are access and identity failures. Enforce least privilege, segmentation, strong identity, and continuous auditing.

   AUTOMATE THE DELIVERY PIPELINE.

   Reliability and speed come from repeatability. CI/CD, tests, migrations, and rollbacks are automated and standard.

   RUN-COST AWARENESS.

   Total cost is dominated by operations and change. Evaluate choices on maintainability, observability, stability, and operational effort.

   PLATFORM-FIRST.

   LEF’s leverage comes from Thinkwise and standardized capabilities. Prefer platform/configuration; custom tech requires ROI and lifecycle ownership; keep UX and API readiness as baseline.

   CONTAINERIZED AND REPRODUCIBLE DEPLOYMENTS WHERE POSSIBLE.

   Reproducibility reduces drift and accelerates recovery. Use containers/immutable patterns when feasible, prefer open standards by default, and document exceptions with migration options.

Next

System archetypes Container archetypes Reference (facts) Run & Support (operate)